Research Steven Latré

Abstract

Brailsports aims to revolutionise endurance training with AI-optimised sports-scientific models, including the Lactate Threshold and Personalized Fitness Fatigue models. The outputs of these models are currently implemented in an online dashboard, providing coaches tailored insights and customized analytics about their athletes' performance, recovery and training needs. This empowers coaches to make data-driven decisions, enhancing the training regimen's effectiveness and efficiency. This project's objective is to elevate our minimal viable product into a commercializable one, targeting the designated beachhead market of endurance coaches. Looking ahead, Brailsports envisions expanding from a coach-centric platform to developing athlete-focused solutions. We envision a sophisticated mobile application that serves as a personal coach in every athlete's pocket. This app will offer dynamic, AI-driven training plans that adjust in real time to reflect the athlete's current performance metrics, physiological responses and evolving objectives. By doing so, Brailsports aims to democratise access to elite-level training insights, making personalized, science-backed training regimens accessible to athletes at all levels. This expansion will empower athletes with the tools they need for optimal performance, fostering a more informed, connected, and personalized approach to endurance training. In addition to the commercialization of the dashboard, we aim to take the first steps in the development of this mobile application in this project.

Researcher(s)

• Promoter: Verdonck Tim
• Co-promoter: Latré Steven

Research team(s)

• Applied mathematics

Project type(s)

• Research Project

Abstract

Due to technological advances, the available amount of data has increased exponentially over the last decade. The field of data science (DS) has followed this growth as it provides an indispensable tool for translating data into insight and knowledge. Where DS was traditionally concerned with learning associations in data, it has become clear in recent times that causal relations often provide a deeper understanding of the data and a stronger tool in many practical applications. One of the established approaches to causal modeling is to use a directed acyclical graph (DAG) to represent the causal relations. These DAGs have to be learned based on observed data. Many of the SOTA techniques for DAG learning are very sensitive to anomalies, and yield unreliable results in their presence. We aim to develop methods for DAG learning that remain efficient and reliable under contamination of the data. The project starts by building a solid foundation for the concepts of robustness in DAG learning. Building upon these foundations, we will then proceed to build a general robust DAG learning methodology. The project envisions three different but complementary approaches to the development of robust DAG learning methods. The developed methodology will be evaluated theoretically and empirically, and tested in a variety of real world cases.

Researcher(s)

• Promoter: Verdonck Tim
• Co-promoter: Latré Steven
• Co-promoter: Raymaekers Jakob
• Fellow: Leyder Sarah

Research team(s)

• Applied mathematics

Project type(s)

• Research Project

Abstract

The European Green Deal relies on healthy forests to remove carbon (C) from the atmosphere, stabilize the water cycle and provide sufficient biomass for the future bioeconomy. The Flanders Forest Living lab realizes a specific breakthrough in the assessment of these crucial ecosystem functions, at spatial scales ranging from the individual tree to the entire forest. The Living Lab is situated in an ICOS flux-tower observatory, that currently already provides a permanent assessment of ecosystem scale CO2-fluxes, evapotranspiration and respiration. To date however, no technique is available to study the function of individual trees, at daily resolution, across a forest. achieving this is the groundbreaking objective of this new infrastructure. Its specific equipment allows for crucial realistic simulation of the water-, energy- and carbon fluxes by advanced vegetation models at spatial scales matching those of satellite imagery products, thereby creating new possibilities for applications such as automated precision forestry management, fire prevention and worldwide carbon budget quantifications. The new infrastructure involves an UAV and a set of linked validation sensors. Observations are steered by artificial intelligence, in order to be able to adapt the flight pattern to the fluctuating source area of the flux-tower, and in order to proactively adapt to specific weather patterns and potentially interesting ground-sensor observations.

Researcher(s)

• Promoter: Janssens Ivan
• Co-promoter: Campioli Matteo
• Co-promoter: Gielen Bert
• Co-promoter: Latré Steven
• Co-promoter: Nijs Ivan
• Co-promoter: Roland Marilyn
• Co-promoter: Scheunders Paul
• Co-promoter: Vicca Sara

Research team(s)

• Plant and Ecosystems (PLECO) - Ecology in a time of change

Project type(s)

• Research Project

Abstract

This project is geared towards discovering and developing new applications of state-of-the-art psychophysiological sensor technologies (using computational and AI techniques) to help people with different needs work, learn and play in our modern society, ensuring that this tracking is meaningfully and responsibly applied. To accomplish this, our consortium is suitably interdisciplinary. This undertaking requires well-controlled lab studies and (near-)continuous psychophysiological tracking in real-life settings 'in the wild'. The research infrastructure applied for enables flexible movement from lab explorations of promising markers to checking their robustness in realistic, ecological contexts, and back again.

Researcher(s)

• Promoter: Gijbels David
• Co-promoter: Daelemans Walter
• Co-promoter: DeSmet Ann
• Co-promoter: Jankowska Anna
• Co-promoter: Latré Steven
• Co-promoter: Poels Karolien
• Co-promoter: Vaes Kristof
• Co-promoter: Van de Cruys Sander
• Co-promoter: Van den Bossche Piet
• Co-promoter: Van Waes Luuk

Research team(s)

• EduBROn

Project type(s)

• Research Project

Abstract

CalcUA stimulates the use of scientific and technical computing by providing access to state-of-the art computer hardware infrastructure. It shares knowledge, expertise, and training on the efficient use of this hardware in combination with the best available algorithms. It makes it possible to solve largescale scientific problems in a distributed way. In this way users will take advantage of the latest possibilities of scientific and technical computing in their research and R&D. It creates an environment for the exchange of ideas and expertise on large-scale simulation and the processing of large sets of data and related scientific problems. It is part of the Flemish Supercomputer Centre, which provides part of the funding for the personnel and hardware. Funding as a core facility will create a multiplier effect at UAntwerpen by investing in training, community building and the creation of new applications and externally funded joint projects between research groups, CalcUA, and local industry at the national and international level.

Researcher(s)

• Promoter: Vanroose Wim
• Co-promoter: Becuwe Stefan
• Co-promoter: De Winter Hans
• Co-promoter: Herrebout Wouter
• Co-promoter: Latré Steven

Research team(s)

• Applied mathematics

Project type(s)

• Research Project

Abstract

Neuromorphic computing is an emerging field of research. In machine learning, spiking neural networks (SNN) are now widely used to exploit the low-power consumption promise of these brain-inspired systems, saving up to an order of magnitude of energy in inference. Recently, advanced training methods for spiking neural networks have been developed to bridge the performance gap with deep learning, enabling use in real life applications at the edge such as continuous heart rate monitoring in smartwatches or on-sensor detection of dangerous sounds. More precisely, the Liquid State Machine (LSM) a recurrent reservoir-based SNN, has come forward as a simple, yet inherently very powerful computational framework for spatio-temporal data processing. The spike-based processing of time-series in a reservoir allows the LSM to retrieve features in a unique way. There are many open research questions, such as what type of learning best suits the neuromorphic reservoir and how multiple reservoirs can be connected in an optimal way so the most important features are passed through. In this proposal we introduce new spike-based learning rules that will allow us to derive relevant features inside the LSM, optimally connect multiple reservoirs by focusing on the important features and consequently boost the performance of LSM at low power consumption.

Researcher(s)

• Promoter: Latré Steven
• Fellow: Deckers Lucas

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The trend towards autonomous drones is currently driving the integration of an increasing number of sensors for safe navigation under all circumstances, forcing algorithms and hardware to be energy efficient and fast. When drone technology continues to mature, deploying swarms of them will enable even more advanced use cases, for example in precision agriculture. Swarms also offer the possibility of sharing both sensory and compute resources, making the swarm act and respond as a single collaborative entity with overall better performance. In this PhD project, we work with real-world multi-sensory data collected by multiple drones and develop a spike-based neuromorphic fusion solution running on custom imec hardware. More specifically, we will focus on the following research questions: - Can we build a low power sensor fusion solution based on spiking neural networks for autonomous drone navigation and obstacle avoidance, running on imec hardware. We will investigate different solutions to perform spike encodings and carry out the learning. A trade off will be made vs power consumption and hardware. - How can collaborative drones, each with their own spike-based neuromorphic fusion solution, communicate with each other in a timely and resource efficient way? Which sensor fusion tasks need to be performed by which nodes in a collaborative setting? - Can we develop efficient techniques for distributed training across multiple spike- based drones to reduce each drone's individual memory and power requirement and, at the same time, lower the convergence time of the swarm?

Researcher(s)

• Promoter: Latré Steven
• Fellow: Van Damme Laurens

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Conventional climate change mitigation alone will not be able to stabilise atmospheric CO2 concentrations at a level compatible with the 2°C warming limit of the Paris Agreement. Safe and scalable negative emission technologies (NETs), which actively remove CO2 from the atmosphere and ensure long-term carbon (C) sequestration, will be needed. Fast progress in NET-development is needed, if NETs are to serve as a risk-hedging mechanism for unexpected geopolitical events and for the transgression of tipping points in the Earth system. Still, no NETs are even on the verge of achieving a substantial contribution to the climate crisis in a sustainable, energy-efficient and cost-effective manner. BAM! develops 'super bio-accelerated mineral weathering' (BAM) as a radical, innovative solution to the NET challenge. While enhanced silicate weathering (ESW) was put forward as a potential NET earlier, we argue that current research focus on either 1/ ex natura carbonation or 2/ slow in natura ecosystem-based ESW, hampers the potential of the technology to provide a substantial contribution to negative emissions within the next two decades. BAM! focuses on an unparalleled reactor effort to maximize biotic weathering stimulation at low resource inputs, and implementation of an automated, rapidlearning process that allows to fast-adopt and improve on critical weathering rate breakthroughs. The direct transformational impact of BAM! lies in its ambition to develop a NET that serves as a climate risk hedging tool on the short term (within 10-20 years). BAM! builds on the natural powers that have triggered dramatic changes in the Earth's weathering environment, embedding them into a novel, reactor-based technology. The ambitious end-result is the development of an indispensable environmental remediation solution, that transforms large industrial CO2 emitters into no-net CO2 emitters.

Researcher(s)

• Promoter: Latré Steven
• Co-promoter: Oramas Mogrovejo José Antonio
• Co-promoter: Verdonck Tim

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The IOF consortium IDLab is composed of academic supervisors at the IDLab Research Group, a UAntwerp research group with members from the Faculty of Science and the Faculty of Applied Engineering. IDLab develops innovative digital solutions in the area of two main research lines: (1) Internet technologies, focusing on wireless networking and Internet of Things (IoT), and (2) Data science, focussing on distributed intelligence and Artificial Intelligence (AI). The mission of the IDLab consortium is to be the number one research and innovation partner in Flanders and leading partner worldwide, in the above research areas, especially applied in a city and its metropolitan surroundings (industry, ports & roads). To realize its mission, IDLab looks at integrated solutions from an application and technology perspective. From an application point of view, we explicitly provide solutions for all stakeholders in metropolitan areas aiming to cross-fertilize these applications. From a technological point of view, our research includes hardware prototyping, connectivity and AI, enabling us to provide a complete integrated solution to our industrial partners from sensor to software. Over the past years, IDLab has been connecting the city and its surroundings with sensors and actuators. It is time to (1) reliably and efficiently connect the data in an integrated way to (2) turn them into knowledgeable insights and intelligent actions. This perfectly matches with our two main research lines that we want to extensively valorise the upcoming years. The IDLab consortium has a unique position in the Flemish eco-system to realize this mission as it is strategically placed across different research and innovation stakeholders: (1) IDLab is a research group embedded in the Strategic Research Centre imec, a leading research institute in the domain of nano-electronics, and more recently through groups such as IDLab, in the domain of digital technology. (2) IDLab has a strategic link with IDLab Ghent, a research group at Ghent University. While each group has its own research activities, we define a common strategy and for the Flemish ecosystem, we are perceived as the leading partner in the research we are performing. (3) IDLab is the co-founder of The Beacon, an Antwerp-based eco-system on innovation where start-ups, scale ups, etc. that work on IoT and AI solutions for the city, logistics, mobility and industry 4.0 come together. (4) Within the valorisation at UAntwerp, IDLab contributes to the valorisation within the domain 'Metropolitanism, Smart City and Mobility'. To realize our valorisation targets, IDLab will define four valorisation programs: VP1: Emerging technologies for next-generation IoT; VP2: Human-like artificial Intelligence; VP3: Learning at the edge; VP4: Deterministic communication networks. Each of these valorisation programs is led by one of the (co-)promoters of the IDLab consortium, and every program is composed of two or three innovation lines. This way, the IDLab research will be translated into a clear program offer towards our (industrial) partners, allowing us to build a tailored offer. Each valorisation program will contribute to the different IOF objectives, but in a differentiated manner. Based on our current experience, some valorisation programs are focusing more on local partners, while others are mainly targeting international and EU funded research projects.

Researcher(s)

• Promoter: Hellinckx Peter
• Promoter: Latré Steven
• Promoter: Mannens Erik
• Promoter: Weyn Maarten
• Co-promoter: Famaey Jeroen
• Co-promoter: Hellinckx Peter
• Co-promoter: Latré Steven
• Co-promoter: Mannens Erik
• Co-promoter: Marquez-Barja Johann
• Co-promoter: Mercelis Siegfried
• Co-promoter: Mets Kevin
• Co-promoter: Oramas Mogrovejo José Antonio
• Co-promoter: Saldien Jelle
• Co-promoter: Verdonck Tim
• Co-promoter: Weyn Maarten
• Fellow: Braem Bart
• Fellow: Braet Olivier

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

This project is devoted for the maintenance of the City of Things Hercules infrastructure . Within this project, we have developed the CityLab testbed which is a wireless edge computing platform for smart cities. This provides experimental access to wireless networking infrastructure, edge computing infrastructure and smart city sensors.

Researcher(s)

• Promoter: Hellinckx Peter
• Promoter: Latré Steven
• Promoter: Mannens Erik

Research team(s)

• Internet Data Lab (IDLab)
• Modelling for Sustainability (M4S)

Project type(s)

• Research Project

Abstract

In this IOF POC, we aim to bring together the expertise in sports science and machine learning to develop intelligent tools for coaching endurance sports. These tools will assist the coach in tracking the fitness level of the athletes and provide early warning of any potential issues within the physiological data. By leveraging the power of machine learning, we hope to create a more efficient and effective coaching process that can help athletes reach their full potential. Additionally, by integrating sports science knowledge, we aim to ensure that the tools we develop are grounded in the latest research and understanding of how the body responds to endurance training.

Researcher(s)

• Promoter: Verdonck Tim
• Co-promoter: Latré Steven

Research team(s)

• Applied mathematics

Project type(s)

• Research Project

Abstract

The Flanders AI Research Program focuses on demand-driven, leading-edge, generic AI research for numerous applications in the health and care sector and industry, for governments and their citizens. The requirements were indicated by users from these application domains.

Researcher(s)

• Promoter: Mannens Erik
• Co-promoter: Calders Toon
• Co-promoter: Daelemans Walter
• Co-promoter: Goethals Bart
• Co-promoter: Latré Steven
• Co-promoter: Laukens Kris
• Co-promoter: Martens David
• Co-promoter: Sijbers Jan
• Co-promoter: Steckel Jan

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

There is a lack of reliable information regarding the European vegetation's current elemental composition and the factors that affect it. We investigate 183928 georeferenced records on woody plant's foliar concentrations of Calcium (Ca) Magnesium (Mg), and Sulfur (S) from published databases, which allow us to create European maps of foliar Ca, Mg, and S foliar concentrations using machine learning at a resolution of 1 km^2 for woody plants. We aim to understand the relationship of ecological processes particularly nutrient cycling in the complex ecosystem analysis. For this, we develop a statistical predictive model of foliar concentration of Ca, Mg, and S.

Researcher(s)

• Promoter: Latré Steven
• Fellow: Narsingh Vaidehi

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

In the emerging field of low power AI for deployment at edge devices, Spiking Neural Networks (SNN) are gaining traction as prime candidate technology due to initial results of spiking neuromorphic systems saving up to one or two order of magnitude in energy for inference tasks. While today's SNNs are typically trained in the cloud using variants of the traditional backpropagation method, future applications will benefit from on-device adaptation and learning capabilities. Spike-Timing-Dependent Plasticity (STDP), an interesting brain-inspired local learning alternative that uses the temporal factor of spike events for learning, has shown promising results for unsupervised feature learning, and can be deployed for on-device learning. However, for training on specific tasks, STDP needs to be extended with a third factor in the form of a success signal to steer the learning process. The existing three-factor learning rules can be characterized by having different and somewhat ad-hoc definitions for the third factor which may or may not work well in particular applications. This proposal will investigate new SNN training methods that combine STDP learning with formal methods from Temporal Logic to define structured reward signals that are applicable to a wide range of supervised, self-supervised and reinforcement learning applications, and allow for distributed deployment. Enhanced SNNs will open up a wealth of opportunities for smart industries, health, environment etc.

Researcher(s)

• Promoter: Latré Steven
• Fellow: Van Damme Laurens

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Changes in climate can greatly affect the phenology of plants, which can have important feedback effects, such as altering the carbon cycle. These phenological feedback effects are often induced by a shift in the start or end dates of the growing season of plants. The normalized difference vegetation index (NDVI) is a simple indicator that can be used to determine whether the area being observed contains green vegetation and can be used to approximate the growing season of plants. In this project, we apply machine learning techniques to investigate the relationship between soil temperature and the NDVI curve in a unique ecosystem in Iceland and to find out whether this relationship is modulated by climatic variables.

Researcher(s)

• Promoter: Latré Steven
• Fellow: Bussmann Bart

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The Flemish AI research program aims to stimulate strategic basic research focusing on AI at the different Flemish universities and knowledge institutes. This research must be applicable and relevant for the Flemish industry. Concretely, 4 grand challenges 1. Help to make complex decisions: focusses on the complex decision-making despite the potential presence of wrongful or missing information in the datasets. 2. Extract and process information at the edge: focusses on the use of AI systems at the edge instead of in the cloud through the integration of software and hardware and the development of algorithms that require less power and other resources. 3. Interact autonomously with other decision-making entities: focusses on the collaboration between different autonomous AI systems. 4. Communicate and collaborate seamlessly with humans: focusses on the natural interaction between humans and AI systems and the development of AI systems that can understand complex environments and can apply human-like reasoning.

Researcher(s)

• Promoter: Hellinckx Peter
• Promoter: Latré Steven
• Co-promoter: Calders Toon
• Co-promoter: Daelemans Walter
• Co-promoter: Goethals Bart
• Co-promoter: Laukens Kris
• Co-promoter: Martens David
• Co-promoter: Sijbers Jan
• Co-promoter: Steckel Jan

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Traditional machine learning techniques focus on developing predictive models that have the sole purpose of obtaining a high degree of accuracy on a given data set. These types of models exploit any type of association between the input and target variables that may increase the performance. However, in practice, the training and test distribution often differ significantly, resulting in unreliable and failing models. The key to learning generalizable models that work in a broad range of environments (and that are not affected by small changes in the test distribution) lies in learning causal predictive features. However, learning causal models under changing environments and in systems with hidden confounders is an unsolved problem and is directly connected to the generalisation gap. In this project, we aim to use the novel framework of causal machine learning to develop algorithms that can handle changing environments. More specifically, this project focuses on learning invariant and causal representations from data using causal machine learning. The results are models that are proven to be more generalizable, can cope with interventions, and are able to extract interpretable causal relations directly from data.

Researcher(s)

• Promoter: Latré Steven
• Co-promoter: Nys Jannes
• Fellow: Mortier Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Deep learning has surely become a buzzword in the last decade, but rightly so: it is an extremely powerful tool that learns from large amounts of past data, which has significantly outperformed the previous state-of-the-art practices in image processing, language translation, speech and object recognition, biomedicine, drug design, etc. It is ubiquitous in our daily realities - Google Translate, Google Maps, Alexa, Siri, our phone's face or fingerprint unlock feature all rely on deep learning, and it will be of crucial importance for self-driving cars. The practical success of deep learning, however, goes far beyond theoretical understanding. How do deep neural networks work and learn? How well will the network generalize to unseen data? When does it fail, and how can this be avoided? My goal is to shed some light on the last question, by trying to identify the classes of problems for which deep learning performs poorly. In particular, we plan to examine some problems where we would expect topological data analysis to outperform the results obtained with deep neural networks. Topology studies shape, and we expect it to be better in detecting the number of connected components, holes and voids in higher dimensions, or shape convexity; but recent results indicate that the same might be true for detecting shape curvature. We plan to investigate this experimentally, by comparing the results obtained with the two approaches, on a number of diverse synthetic datasets and data available in the literature. In addition, deep learning is expected to underperform when there is not a lot of data available, or when the data is noisy - we will therefore also include such scenarios in our computational experiments. Topological features can thus be recommended as an alternative to deep learning whenever they promise a superior performance, but the findings will also provide us with inspiration on how to improve existing deep architectures, with, for example, an additional network layer for topological signatures, or topological loss functions for network's prediction error.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Update and maintain the AI model every quarter. imec project on recycling electronic waste automatically instead of manually by using an AI model. The AI model was developed in a previous project and this project is for the maintenance of the model and making the necessary updates.

Researcher(s)

• Promoter: Latré Steven
• Co-promoter: Mercelis Siegfried

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Framework agreement to submit to research projects of the Agentschap Telecom (Netherlands). Possible topics are : - Dynamic Spectrum Management & Sharing - Exploration of the roles of the Netherlands Radiocommunications Agency in the energy transition - etc.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The project targets (i) a novel, virtual and distributed data ecosystem for food delivery to physical shops that becomes hyper responsive and efficient thanks to (ii) more accurate forecasting and personalization models that use enhanced AI and scheduling technologies to (iii) optimize the end-to-end logistics from farmers to Distribution Centers (DCs) to stores and customers.

Researcher(s)

• Promoter: Latré Steven
• Co-promoter: Hellinckx Peter
• Co-promoter: Verdonck Tim

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The main objective of the 'Smart Port' COOCK project is to improve operational efficiency within the context of the port, by applying intelligent technologies, aimed at SMEs, by increasing digital maturity through data-driven digitization. The proposed COOCK focuses on two target groups: the value chain within a port context, such as terminal operators, skippers, agents, transport companies, forwarders, shipping companies, rail operators, port authorities, ... and also technology integrators: start-ups, scale-ups, IT- companies, etc. that are active in implementation processes in a port environment.

Researcher(s)

• Promoter: Latré Steven
• Co-promoter: Hellinckx Peter

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

DAEMON - The success of Beyond 5G (B5G) systems will largely depend on the quality of the Network Intelligence (NI) that will fully automate network management. Artificial Intelligence (AI) models are commonly regarded as the cornerstone for NI design; indeed, AI models have proven extremely successful at solving hard problems that require inferring complex relationships from entangled and massive (e.g., traffic) data. However, AI is not the best solution for every NI task; and, when it is, the dominating trend of plugging 'vanilla' AI into network controllers and orchestrators is not a sensible choice. Departing from the current hype around AI, DAEMON will set forth a pragmatic approach to NI design. The project will carry out a systematic analysis of which NI tasks are appropriately solved with AI models, providing a solid set of guidelines for the use of machine learning in network functions. For those problems where AI is a suitable tool, DAEMON will design tailored AI models that respond to the specific needs of network functions, taking advantage of the most recent advances in machine learning. Building on these models, DAEMON will design an end-to-end NInative architecture for B5G that fully coordinates NI-assisted functionalities. The advances to NI devised by DAEMON will be applied in practical network settings to: (i) deliver extremely high performance while making an efficient use of the underlying radio and computational resources; (ii) reduce the energy footprint of mobile networks; and (iii) provide extremely high reliability beyond that of 5G systems. To achieve this, DAEMON will design practical algorithms for eight concrete NI-assisted functionalities, carefully selected to achieve the objectives above. The performance of the DAEMON algorithms will be evaluated in real-world conditions via four experimental sites, and at scale with data-driven approaches based on two nationwide traffic measurement datasets, against nine ambitious yet feasible KPI targets.

Researcher(s)

• Promoter: Marquez-Barja Johann
• Co-promoter: Camelo Botero Miguel
• Co-promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project website

Project type(s)

• Research Project

Abstract

In manufacturing solutions, it is important to quickly educate the operators for their specific task at hand. In this project, the goals is to develop an ergonomic model (digital twin) of a workcell (with tools as robots, equipment, tools). This consists of several aspects: - VR enabled assessment -> 'personalized' training (learning and automated examination by assessing behaviour)​ - FOCUS on balance between: ergonomics, well-being and productivity​ - Deriving mental load and ergonomics factors for advise on improved workcell layout​ ​

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

COVID stressed the need for accurate monitoring of crowds (including density estimation). However, the market solutions currently lacs privacy by design, accuracy, and applicability.​ Our solution is to automatically monitor the crowd by​ combining radar and sensing of ubiquitous radio frequency communication.​ Like the visual spectrum, the radio spectrum is constantly lit up by energy sources.​ Reusing these sources is efficient, and yet really hard to identify people afterwards.​ What we need to prove, is that it can be accurate.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

In the pursuit of developing an AI chip that can efficiently support a quadrillion parameter neural network model it is of utmost importance to define the right AI applications and their AI capabilities that can drive this chip. In this project, the goal is to have an early detection of future applications and required AI capabilities, to identify possible algorithmic approaches, and to allow dedicated technology projects as follow-up. For this, we do a systematic review of future AI evolutions and aim at defining a specific moonshot and roadmap to follow

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

This project originates from the known interest of the Port of Antwerp in finding an innovative and cost-effective way to detect and prevent overloading of the bollards since the quay walls need to be able to cope with the ever-increasing loads and operational times. Overloaded bollards can become a danger to port operations, resulting in e.g. ship ropes that come loose and vessels hitting the docks or bollards that are pulled from the quay towards the vessels. The port initially searched for existing solutions on the market but rapidly realized that there was not such a "market-ready" solution available to tackle this challenge and therefore launched a call for proposals. ID lab proposed an intelligent early-warning system of interconnected bollards to monitor port infrastructure solution, which went beyond expectations in terms of operational value, by combining two of our expertise domains (i) sensors and energy efficient wireless communication protocol and (ii) intelligent data processing with machine learning techniques. While the solution holds great potential some research activities are still needed to test the set-up of such a system in an operational environment. We have state of the art sensing technology that can be applied everywhere but the anomaly detection methods demand fine-tuning on real data from the operational environment. Within this POC we will start a trajectory to gain insights into the breadth of the use and market potential for the technology and get an accurate picture of the technological and contextual requirements that can enhance its adoption by a technology provider that scales it up to the port. Market niches opportunities in the short and long term will be identified by working in close collaboration with the port and its chainport network (e.g. Ports of Rotterdam, Hamburg and LA).

Researcher(s)

• Promoter: Hellinckx Peter
• Co-promoter: Latré Steven
• Co-promoter: Weyn Maarten

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Data relevant for passerby scanning is collected in silos and/or with a specific use case in mind. There is no technical solution for merging that data nor for assessing mobility comprehensively. The current state of the art in this domain is that - Data brokers exist, unlocking urban data and presenting it without further processing. - Data providers recognize the limits of the data they collect. - Individual partners have set up (2-way) collaborations with limited success. - Fusion of available data for one mobility view is not possible at this point. - Silo solutions (verticals) exist. - Academic studies using GPS traces abound, but are not acceptable in a GDPR-governed B2B market. In this project, we develop a framework for establishing a mobility data economy, which serves as a neutral playing field for brokering available mobility data. The goal is that it is fused for accurate assessments and as input for the next generation of mobility models. We take into account privacy & ethics, performance, validation of data and the needs of policy. The CityFlows platform, datasets, consortiums and stakeholders are essential components for enabling this novel business ecosystem.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Reinforcement learning is an active field in machine learning where an agent learns to perform a task by interacting with the environment and receiving positive or negative rewards depending on the chosen actions. Recently, reinforcement learning has seen some big breakthroughs in beating the best human players various tasks, such as the classic board game Go and the popular video game StarCraft II. One of the reasons why the architectures that were used are so successful is that deep learning modules are used which can perform some form of relational reasoning. This allows them to view the environment in terms of distinct objects and make use of the relations between these objects. The field of relational reinforcement learning looks at how these relations between objects can be learned and used to optimally solve the given tasks. As this field is relatively new, there are still many open research questions, such as how to best create a representation of the environment based on these objects and relations and how to improve the efficiency of these networks by learning only the important relations while ignoring the irrelevant ones. In this proposal we introduce new relational reinforcement learning architectures that will allow us to efficiently represent the environment in a relational way, improve efficiency by focusing on the important relations in this representation and increase the ability to generalize to unseen tasks.

Researcher(s)

• Promoter: Latré Steven
• Fellow: Avé Thomas

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Venue mapping is a special case of the reverse geocoding problem. Given user's GPS coordinates, an accuracy radius and a list of venues located inside that radius, we want to derive which venue did the user visit. Unfortunately, noise in the signal, and especially in dense urban areas, limits our ability to achieve satisfactory results. Resent research shows that it is possible to improve the results by incorporating temporal and behavioral knowledge into the venue mapping model. As a company specializing in analyzing sensor data, such as accelerometer, gyroscope and GPS, from mobile devices, Sentiance has a vast amount of data for thousands of users. An open question is how to represent the data so that the model could be trained in fully data-driven fashion. Manually creating rules or labelling millions of venues is not an option and would not result in a scalable, future-proof solution. Restricted by the lack of labelled data, we studied the latest achievements in Deep self-supervised learning in order to design a model that would be able to autonomously reveal the internal patterns available in the unlabeled data. In order to guarantee rich generalization capabilities of our model, we searched for ways to incorporate more knowledge into our model by means of publicly available data and Transfer learning. Despite the fact that such datasets exist, we faced another problem – the format of the data is so different from our in-house data, that none of the existing Transfer Learning techniques could be applied directly. Finally, to tackle this challenge we studied the fields of Multimodal learning and Multi-task learning. In this project we propose training a series of Deep learning models with a novel architecture that would result in a new state-of-the-art solution for the venue mapping problem.

Researcher(s)

• Promoter: Hellinckx Peter
• Co-promoter: Latré Steven
• Fellow: Musaev Gadzhi

Research team(s)

• Modelling for Sustainability (M4S)
• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Inland Waterborne Transport (IWT) advantages as low-energy and low CO2 emitting transport mode are not fully exploited today due to gaps in the logistics system. Inland container vessels pay 6-8 calls at seaport terminals with long waiting times. More time is lost by sub-optimal navigation on rivers and waiting at bridges and locks. In addition, low load factors of containers and vessels impact the logistics systems with unnecessary high numbers of containers being transported and trips being made. NOVIMOVE strategy is to "condense" the logistics system by improving container load factors and by reducing waiting times in seaports, by improved river voyage planning and execution, and by facilitating smooth passages through bridges and locks. NOVIMOVE's innovations are: (1) cargo reconstruction to raise container load factors, (2) mobile terminals feeding inland barges, (3) smart river navigation by merging satellite (Galileo) and real time river water depths data, (4) smooth passage through bridges/locks by dynamic scheduling system for better corridor management along the TEN-T Rhine-Alpine (R-A) route, (5) concepts for innovative vessels that can adapt to low water condition while maintaining a full payload, and (6) close cooperation with logistic stakeholders, ports and water authorities along the R-A route: Antwerp, Rotterdam, Duisburg, Basel. NOVIMOVE technology developments will be demonstrated by virtual simulation, scaled model tests and full-scale demonstrations. NOVIMOVE innovations will impact the quantity of freight moved by IWT along the R-A corridor by 30% with respect to 2010 baseline data. The NOVIMOVE 21-members consortium combines logistics operators, ports, system-developers and research organisations from 4 EU member states and two associate countries. The work plan contains 4 technical Work Packages. The project duration is four years; the requested funding is 8,9 MIO.

Researcher(s)

• Promoter: Hellinckx Peter
• Co-promoter: Latré Steven

Research team(s)

• Modelling for Sustainability (M4S)
• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Recycl.AI is studying the use of deep learning in the circular economy. This project will start from an existing deep-learning based computer vision algorithm, which has been developed in the past by IDLab. The algorithm is able to predict the product category of e-waste.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Interaction between humans and artificial intelligence (AI) is still lacking the degree of engagement and entrainment that characterizes interaction between humans. The project WithMe aims at bridging this gap by investigating in detail the processes that happen in the human brain when engaged in an activity together with an other person: pursuing a common goal or simply enjoying a common activity. The brain signals that will be explored are characteristic for attention, emotion, and reward. Based on findings with people, WithMe will investigate the characteristics of an audiovisual interaction that makes it feel human, but the project will also explore whether an AI system could benefit from direct access to biomonitoring of the person it is communicating with. The new human-AI approach thus derived will open a wealth of new applications in health, revalidation, communication and information sharing, entertainment, etc.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Open city - Knowledge transfer with respect to the disclosure of standards and best practices for an open 'datafied' society within the domain of environmental parameters (air quality, sound, water, light, electromagnetic exposure...) as a foundation for innovative value models.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Today, drones are sophisticated observers. They can capture data more efficiently than traditional alternatives. They can also significantly reduce risks associated with specific observations, eliminating the need for humans to be physically present in hazardous environments. The drones of tomorrow will evolve from mere observers to highly automated, autonomously operating and even decision-making tools. The sky's the limit for the applied science of flying robots – or "dronebots". in this project, we investigate how a low power radar can be integrated with low power neural network models to perform accurate obstacle avoidance and drone navigation.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

By making waterway transport more cost-efficient, Smart Waterway will enable a modal shift for last mile urban logistics from the road to the small waterways in many European cities, including a city as Ghent. For small barges that could enter these waterways, however, the cost of automating a vessel is high compared to the construction cost. Hence, a cost reduction in automating small vessels will be crucial in this shift. We believe this can only be reached by drastically reducing the equipment cost on the autonomous vessel. This does, however, require significant advances in sensing and localization technology. Although a lower accuracy is sufficient for autonomous waypoint-based navigation, low-cost onboard sensors will not suffice in more complex scenarios (i.e. locks, bottlenecks such as bridges, loading and unloading bays) where accurate localization is needed to safely maneuver the vessel. To overcome this issue, these critical locations will be equipped with additional sensors (e.g., IR, cameras) and a novel ultra-wideband localization system. By combining low-cost onboard sensors with infrastructure near critical locations, Smart Waterway aims to achieve economically viable level 3 autonomy in urban waterways.

Researcher(s)

• Promoter: Hellinckx Peter
• Co-promoter: Latré Steven
• Co-promoter: Marquez-Barja Johann

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Emerging artificial intelligence applications in professional sports based on the use of sensors, wearables and video data offer huge opportunities to innovate live sports reporting, but the code to turn these data sources into attractive and meaningful stories has not yet been cracked. To unleash this potential, DAIQUIRI will develop a media-focused sensor data platform and professional dashboard allowing content creators to augment live sports experiences. DAIQUIRI targets both real-time augmentation of live TV and near-live story snippet inserts in an interactive set-top-box application layer. The project demonstrator will showcase end-to-end sensor data integration for reporting of hockey and cyclocross.  DAIQUIRI will address the current challenges of data tsunami handling,  sensor-video matching and enrichment, insights-driven captation and enable the automated orchestration of multi-modal story snippets through editorial AI algorithms. The consortium covers the full value chain bringing together unique expertise in sports event capturing (Videohouse, NEP), enriched sensor data platform (InTheRace, Arinti, imec-IDLab), editorial tooling and storytelling (VRT) and interactive user experiences in a living room environment (Telenet, imec-MICT).

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The goal of the HAI-SCS (Helicus Aero Initiative – Scheduling Connectivity Security) project is to enable complete, secure and safe automation of mission-critical UAV flights focusing on medical transport. To enable the above described objectives, important new technological innovations are planned within the HAI consortium SCS project: * An automated flight planning and scheduling algorithm able to learn in real time the best flight plan and schedule, given a high dimensional set of input parameters and multi-modal output options (flights, ground transport). Given the high dimensionality of the problem, the aim is to reinforce a learning-based approach, where the total reward of all UAS flights is maximized. A phased approach is being proposed in which a fixed corridor-based airspace design is being assumed in a first phase, allowing an operational handshake-based flight approval process that can be set up with the authorities. The second phase involves the inclusion of a flexible airspace design model in the flight planning and scheduling process. * A dynamic heterogenous quality of service (QoS) management layer able to provide seamless QoS across multiple communication channels (e.g., 5G, 4G, private networks, direct C-band links, etc.). Moreover, guaranteed connectivity, meeting the QoS requirements, need to be constructed and scaled up/down instantly. * A versatile security management system that provides the building blocks to secure the communication and control of the UAS taking into account the specific security and performance requirements of the application (e.g., low latency, high bandwidth) and the resource constraints of the UAS (e.g., battery capacity).

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The Flemish AI research program aims to stimulate strategic basic research focusing on AI at the different Flemish universities and knowledge institutes. This research must be applicable and relevant for the Flemish industry. Concretely, 4 grand challenges 1. Help to make complex decisions: focusses on the complex decision-making despite the potential presence of wrongful or missing information in the datasets. 2. Extract and process information at the edge: focusses on the use of AI systems at the edge instead of in the cloud through the integration of software and hardware and the development of algorithms that require less power and other resources. 3. Interact autonomously with other decision-making entities: focusses on the collaboration between different autonomous AI systems. 4. Communicate and collaborate seamlessly with humans: focusses on the natural interaction between humans and AI systems and the development of AI systems that can understand complex environments and can apply human-like reasoning.

Researcher(s)

• Promoter: Latré Steven
• Co-promoter: Calders Toon
• Co-promoter: Daelemans Walter
• Co-promoter: Goethals Bart
• Co-promoter: Hellinckx Peter
• Co-promoter: Laukens Kris
• Co-promoter: Martens David
• Co-promoter: Sijbers Jan
• Co-promoter: Steckel Jan

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Climate change will affect Arctic ecosystems more than any other ecosystem worldwide, with temperature increases expected up to 4-6°C. While this is threatening the integrity and biodiversity of the ecosystems in itself, the larger ecosystem feedbacks triggered by this change are even more worrisome. During millions of years, atmospheric carbon has been stored in the Arctic soils. With warming, the carbon can rapidly escape the soils in the form of CO2 and (even worse) the strong greenhouse agent CH4. Despite decades of research, scientists still struggle to unveil the scale of this carbon exchange, and especially how it will interact with climate change. An overarching question remains: how much carbon will potentially escape the Arctic in the future climate, and how will this affect climate change? FutureArctic embeds this research challenge directly in an inter-sectoral training initiative for early stage researchers, that aims to form "ecosystem-of-things" scientists and engineers at the ForHot site. The FORHOT site in Iceland offers a geothermally controlled soil temperature warming gradient, to study how Arctic ecosystem processes are affected by temperature increases as expected through climate change.

Researcher(s)

• Promoter: Latré Steven
• Co-promoter: Verdonck Tim

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

From 17 October 2019 to 19 April 2020, Kunstenstad is organizing the Cool Japan exhibition on the third floor of the MAS / Museum aan de Stroom. As part of the exhibition, the ""humanoid"" robot Pepper, owned by IDLab, will be on display. The intention is that he performs his humanoid functions and interacts live with visitors to the exhibition (young and old). IDLab rents out Pepper and will program it to start a conversation with visitors, host a quiz about the exhibition and do a dance.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Bikonnector is a proof-of-conceptproject. In this research project, we will focus on connected bikes for a better experience. The existing technology will be further examined with valorization as purpose.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Behavioral economics provides a relevant theoretical framework that can explain and predict individuals' seemingly irrational choices with respect to their health. By understanding individuals as non-rational actors with predictable biases, individuals can be guided or "nudged" toward wiser choices without restricting their choice freedom or significantly changing their economic incentives. The current project focuses on two complementary health risk behaviors, (un)healthy eating and physical (in)activity, and analyses how persuasive cues or 'nudges' can be applied to interactive, 'just-in-time' (JIT) interventions that are adapted to an individuals' unique characteristics, needs and context. To date, a major gap still exists between the technological capacity to deliver adaptive communications, existing theoretical behavioral frameworks and current applications. The main goal of the project is to close this gap by developing and testing an integrative theoretical framework on how just-in-time adaptive interventions affect individuals' health risk behaviors, by (i) adopting an interdisciplinary perspective, (ii) developing the main technological building blocks that enable these JIT adaptive interventions and (iii) testing the effectiveness of different interventions for different individuals in different contexts.

Researcher(s)

• Promoter: Dens Nathalie
• Co-promoter: De Pelsmacker Patrick
• Co-promoter: Latré Steven
• Co-promoter: Poels Karolien
• Co-promoter: Vandebosch Heidi

Research team(s)

• Marketing

Project type(s)

• Research Project

Abstract

Reinforcement Learning agents have attained incredible achievements over the past few years, with AlphaGo's resounding victory over one of the world's top Go players as a crowning achievement. A severe limitation of such agents is that they only know how to function in one very specific environment; AlphaGo is unable to play Go with a tweaked ruleset, let alone play competitively in a different board game. The meta-learning principle aims to improve this. By training the agent not only on one task, but instead on many tasks from a distribution, the trained agent can quickly learn how to behave in a novel task from the distribution. In this project, we propose several improvements to the field of meta-reinforcement learning. First, we propose a meta-learner based on Hierarchical Temporal Memory, which mimics the human brain according to our current understanding of it. This system adapts quickly to changing patterns in the environment—a desirable property for a meta-learner. We also investigate a plethora of ways to auto-generate these task distributions, and evaluate how we can introduce new abilities efficiently to an already trained meta-learner. Finally, we will extend a meta-learner to work with not just one, but with many task distributions. Ideally, such a system would be able to quickly learn to perform any conceivable task at least as well as a human.

Researcher(s)

• Promoter: Latré Steven
• Fellow: Struye Jakob

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Everyone that makes use of wireless communication technologies such as Wi-Fi has definitely already experienced a badly performing network. Simply googling "better Wi-Fi performance" yields 64.000 results! Quite often this resulted in dissatisfaction and frustration, not only because of the bad performance itself, but also the inability to pinpoint the cause of the problem. Why aren't wireless networks sufficiently intelligent to optimise their configuration to the needs of the diverse applications on top? Looking at the latest evolutions, we see that these networks become increasingly flexible, exposing control capabilities that can be used to define how data must be handled. So, there is flexibility and the networks can be managed, but we argue that without rethinking the way how applications fit into network management, one will continue to perform configurations based on incomplete information. This way, suboptimal wireless network performance and user dissatisfaction will remain commonplace. Therefore, Ctrl-APP aims to establish a new networking paradigm, called appdaptive networking. This is achieved by extending the separation of data and control plane, a typical networking concept, to applications. This way, applications become able to pass intentions to the network, the networks can be properly instrumented to perform fine-grained diagnostics and the resulting knowledge can used to automatically learn and enforce the best configuration. - 1

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Rare events in surveillance video streams are defined as events deviating from "normal" operation, and worth reporting or storing for later analysis. They can include cars driving unusually fast, fights on the street, people running away, unidentified objects, … Current state-of-the-art deep neural networks have achieved remarkable successes in terms of accuracy in a variety of AI-related tasks, but typically require a large set of labeled data to train to recognize specific objects and situations in a supervised way. This projects aims at investigating the use of deep neural networks to automatically and unsupervised learn the important features of a camera feed, and later report unusual events in the stream. IDLab-UA will investigate the applicability of 3rd generation neural networks (spiking neural networks) for this use case, and provide imec with insights on the potential of spiking neural networks (in comparison with quantized 2nd generation artificial neural networks). As an important aspect is the possibility of future on-chip learning the focus of the spiking neural networks research track is on using very simple models for spiking neurons, that would ease later implementation in dedicated hardware. The envisaged track is using temporal coding and one spike operational mode in a multi-layer (convolutional) network architecture. Training should be done using only local learning rules, and will start from standard STDP (Spike Timing Dependent Plasticity) and its variants. Further, specialized spiking response models, optimized for embedded inference with short term memory may be developed, and network size reductions techniques explored. The work on spiking neural networks will start from simple image classification (on the MNIST benchmark), provide an initial analysis of the use of dynamic vision sensor cameras (DVS cameras) for motion anomaly detection in video, and conclude with object and motion detection in short video fragments. Results will be compared with a research track using 2nd generation neural networks, executed by IDlab-Gent, and on the same data sets. Evaluation criteria will include the accuracy of the designed algorithms, the ability to reduce the number of false positive detections over time, the (possible) hardware footprint of the designed algorithms, and insights in the overall potential of spiking neural networks on data sets beyond the traditional simple benchmarks.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

imec will develop a detailed device classification model that extends the PoC sub-category classification model. They will then develop a device size classifier model. The next step in the project is the system development and integration, data engineering and reporting. This concludes phase 1 of the projects. In Phase 2, imec will develop a device classiciation model, detailed per device type.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Developing a sub-subcategory classification model that extends the PoC sub-category classification model. (i) Exploratory dataset analysis and dataset cleanup. (ii) Train new deep learning image classification models at sub-sub-category level (e.g., 02.01.a). (iii) Report on model performance.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Artificial Intelligence in general and deep learning specifically has experienced major breakthroughs in the last decade showing above human intelligence for complex tasks. Current deep learning technologies are however very power hungry and therefore require the use of large Graphical Processing Units (GPUs) farms in large-scale datacenters. With the recent advances in neural network hardware (neuromorphic computing design such as Neural Processing Units) we can expect more and more local neural networks being pushed to the edge of the network. However, what is lacking is exploiting the value of networking, and the Internet in particular, by connecting multiple heterogeneous learning systems together and allow more powerful learning-enabled applications to be built on top. The goal of this project is to create a layer which is able to connect multiple heterogeneous learning systems across the Internet so that they can act as a single deep learning system performing both on-line learning and inferencing. For this, we will develop both a low overhead communication protocol and Software Defined Networking-based control layer, which can define how and when different learning systems need to be connected. Finally, we will focus on the adaptation of various learning algorithms to this connected environment so that they are able to easily transfer knowledge from one learning system to the other.

Researcher(s)

• Promoter: Latré Steven
• Fellow: Hutsebaut-Buysse Matthias
• Fellow: Rocco Rodolfo

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Health care is an essential service that uses a great deal of sensitive personal data which has a high black market value being a lucrative target for data theft and ransomware attacks.The EU NIS Directive (EU 2016/1148) and GDPR (EU 2016/679) will harmonize and improve information security in Europe. Both require relevant ICT infrastructure operators to perform risk assessments, introduce appropriate security measures to manage identified risks, and report security breaches. Unfortunately, risk-based approaches are notoriously difficult to implement in a consistent and comprehensive fashion. They depend on a high level of understanding of both cybersecurity and of the system or network to be protected, are labour intensive and costly and typically done by small teams. This is increasingly inappropriate as health care providers introduce IoT systems, cloud services and (in the near future) 5G networks to provide services in which patients are more engaged, may own some of the devices used, and want access in hospitals, on the move or at home. The ProTego project will develop a toolkit and guidelines to help health care systems users address cybersecurity risks in this new environment by introducing 3 main advances over current approaches: Extensive use of machine intelligence: a combination of machine inference exploiting a priory knowledge for security-by-design, and machine learning from data for run-time threat detection and diagnosis; Advanced data protection measures: advanced encryption techniques and hardware based full memory encryption, and multi-stakeholder IAM to control access to and by user devices, to protect data at rest and provide ultra-secure data exchange portals; Innovative protocols for stakeholder education: using security-by-design analysis to target training and support stakeholders to contribute to network overall security.The toolkit will be integrated and validated in IoT and BYOD-based case studies at two hospitals.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The main objective of the Flexnet project is to build a new paradigm for flexible network communication and thus promote IoT value creation. The Flexible IoT network provides IoT value creators the opportunity to consume network communication on demand in real time, automatically and according to specific needs. On a European level the Flexnet Celtic-plus consortium consists of 14 partners from 5 different countries. The Belgian consortium wishes to focus on maintaining the mission-critical communication during emergency situations, by supporting a flexible and reliable communication network. The connectivity between different heterogeneous wireless networks will be orchestrated by a new platform (imec) that integrates all these technologies in a single, cirtual and highly configurable network infrastructure. Micro-services will be rolled out autonomously and migrated live, based on monitored IoT context and application requirements.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Antwerp is the economic stronghold of Flanders, thanks to its world port, the second largest chemical cluster, a strong creative sector, but above all a joint ambition to grow the region through innovation. The university's role has broadened to become a driving force for innovation, an innovative regional force in knowledge-intensive ecosystems. On the one hand by delivering well-trained people; on the other hand, by responding to the specific needs of the innovation ecosystem in which the university is embedded. Based on the strengths of the University of Antwerp combined with the characteristics of the Antwerp ecosystem, the university focuses on three valorisation domains. A pre-incubation structure is being set up within each of these domains to support and strengthen the valorisation processes within these different domains. This is an open innovation hub where physically the different actors of the relevant innovation ecosystem can meet to work together on innovation projects and to follow training programs. One of these valorisation domains set up is metropolitan areas, smart city, mobility and logistics. In this domain, collaborations on IoT and AI projects with applications in smart city, smart industry, smart port and logistics, smart mobility and smart building are set up from The Beacon. The Beacon is the result of a unique collaboration between the city, the port, the university, IMEC, Lantis and Agoria.

Researcher(s)

• Promoter: Latré Steven
• Fellow: Bracke Ilse

Research team(s)

• Internet Data Lab (IDLab)

Project website

Project type(s)

• Research Project

Abstract

This project wants to Perform a first initial study to scope the work and derisk the project to investigate whether Al can help in automating the sampling, design a neural network based approach to automatically classify the electronic devices category and validate this on the existing Recupel labelled dataset; heikkust

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

This project is part of the IMEC Frame Agreement and is being given as structural investment for fundamental research based on yearly set KPIs from the group to IMEC. This A-budget is defined within the IMEC Way of Working and part of the frame agreement of the University of Antwerp and IMEC.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Research on performance optimization of the physical layer of communication networks has been focused on the development of transmission techniques such as MIMO and OFDM/DMT that effectively exploit the space and frequency dimension. However, the physical layer is usually configured statically and thus fails to properly exploit the time dimension. Upper layers in the protocol stack hold crucial information on the time dependent nature of the network traffic, which can indeed be exploited by the physical layer to dynamically select different configurations and increase overall network efficiency. Therefore, the aim of this project is to develop real-time adaptive physical layer control algrithms that can be combined with existing upper layer network functions so as to additionally exploit the time dimension optimally. The possibilities and performance gains of real-time adaptive physical layer control will be explored in the context of fifth generation broadband copper access networks (5GBB). 5GBB envisages a hybrid fiber-DSL deployment where the fiber network is terminated near the boundary between public and private property. In such deployment, the reduced copper loop length, together with the use of specific (e.g. full duplex) transmission techniques, enables data rates of up to 10 Gbps. In addition, the specificity of the deployment scenario is to make, for the first time, the implementation of real-time adaptive physical layer control algorithms a feasible objective.

Researcher(s)

• Promoter: Blondia Chris
• Co-promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Within the Smart Highway project, MOW and imec will build a high tech test environment to support automated driving along (a part of) a highway [10-20 km E313 and a part of the ring of Antwerp R01] combined with a regional road [Turnhoutsebaan N12 towards the city centre of Antwerp]. These roads will be equipped with wireless communication (both 802.11p and LTE-V) and sensor technologies, and concrete test cases will be set up to test these technologies for supporting connected and automated vehicles, based on real-life monitoring and analysis. Besides imec and MOW, also KU Leuven and Flanders Make will contribute to the project.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

In the City of Things initiative, imec, the city of Antwerp and the Flemish Region are working together to make Antwerp a large-scale testbed for the testing and development of smart city technology. With this unique project we want to become a driving force for research into - and the development of - smart cities.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The Cognitive Wireless project aims at developing an AI layer which is able to quickly detect both the technology of other nodes in the network as the type of traffic purely based on IQ samples. For this, a deep learning framework will be developed which is able to classify from the pure IQ samples to a series of pre-determined technologies (WiFi, Bluetooth, etc.) and traffic classes (bursty traffic, narrowband traffic, etc.).

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The Better Than Wired project aims at providing deterministic access to wireless networking so that the same level of Quality of Service guarantees can be given to wireless networks as wired networks feature, with an additional benefit of an increased flexibility. The project will mainly evaluate an industry 4.0 scenario where a plant wants to optimise their connectivity. By combining existing building blocks on programmable network management (e,.g., ORCHESTRA, real-time SDR),the project will feature a high level of flexibility in managing the wireless network. As such, it is possible to quickly anticipate to changes in performance.

Researcher(s)

• Promoter: Latré Steven
• Co-promoter: Marquez-Barja Johann

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The RAMSES consortium is working toward commercialising the EmergencyEye, a system designed to allow ambulance dispatchers to have better communication with patients and bystanders in the field. The innovation, which supports emergency dispatchers with speedy geolocation, diagnosis and audio-visual guidance of resuscitation measures, was tested in a pilot in Rhein-Kreis-Neuss, Germany.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

CONCORDA contributes to the preparation of the European motorways for automated driving and high density truck platooning with adequate connected services and technologies. The main objective of the Action is to assess performances (reliability/availability) of hybrid communication systems, combining 802.11p and LTE connectivity, under real traffic situations. The study prepares also the improvement of the localisation services. As part of the project a validation and demonstration of the developed methods for self-organisation is carried out through extensive simulation experiments, assessing the achievable cost reductions and performance enhancements.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The goal of the FUSE imec.icon project is to do research on and prototype the technical enablers for a flexible federatable service platform allowing to develop, run and manage unified micro-services.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Observer aims to make the traffic lights in the city so smart that they can also "think around" unforeseen problems. We gather information about moving and stationary vehicles using cameras and traffic counters, about how many vulnerable road users there are, about any priority vehicles on the road, etc. We then combine this data with known statistics about normal and abnormal traffic volumes so that we can create a clear picture of how the traffic should be when all of these variables are taken into account. Should some traffic lights stay longer on green? Or shorter? Should the speed limit be adjusted? Day and night, without anyone having to keep an eye on things.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

How much fine dust and electromagnetic radiation is there in the air? The most fundamental ingredient for building a smart city is having access to accurate, relevant and real-time data. What is the air quality in this or that street? How much nitrogen and ozone is there in the air at that intersection? Is the concentration of fine dust the same all over this district? And what's the average particle size? The more we measure, the more we can calculate, forecast and use the data in handy applications: an app that calculates healthy cycle routes, a tool that moves traffic in the right direction, fact-based changes that can be used when a road is being rebuilt, etc. The Dencity project aims to show how big the impact will be on the relevant data we have available – and whether we need to add even more sensors to the city than there are today. We're also looking to see if all of these sensors truly have to be top-quality and which cheaper instruments are available for supplying data of sufficiently high quality.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The Flooding project combines real-time information from sensors in drains and streams, using information from radar images and other useful weather-related input. This information can then be translated into usable information. For instance, the fire brigade knows where best to deploy its manpower, what precautions need to be taken and whether neighboring brigades and resources will have to be brought in. Residents can also be properly informed, with warnings telling them exactly when the risk times will be and what action they need to take. The managers of the drainage system can also be quicker in locating where problems are likely to occur in the network, which goes hand in hand with proactive maintenance, etc.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The IoT domain is characterized by many applications that require low-bandwidth communications over a long range, at a low cost and at low power. This has given rise to novel 'SIM-less' radio technologies that try to fill in this existing market gap of low-power wide area IoT networks, often referred to as Low Power Wide Area Networks or LPWANs. Due to the use of sub-GHz radio frequencies (typically 433 or 868 MHz), a single LPWAN base station has a large coverage area, with typical transmission ranges in the order of 1 up to 50 kilometres. As a result, a single base station can support high numbers of connected devices (> 1000 per base station), allowing a broad range of new technology companies to easily enter the IoT market. Currently, several sub-GHz technologies are being promoted simultaneously, all of which use the same (limited) wireless spectrum. Notorious initiatives in this domain are LoRa, SigFox, IEEE802.15.4g and the upcoming IEEE 802.11ah (or "HaloW") standard. However, many of these technologies are still in their infancy, and optimizations regarding a.o. quality of service, roaming, and service management are still lacking. In addition, since the amount of available spectrum is much smaller and the propagation ranges much larger, these technologies will cause interference at much larger scale, leading to severe inter-technology and inter-operator interference. If left unchecked the unlicensed sub-1GHz bands will soon be congested and unreliable. To avoid this fate, the IDEAL-IoT project will design and develop advanced, highly configurable networking components, combined with a coordination framework to uniformly manage and optimize an ecosystem of coexisting wireless sub-GHz LPWANs. More specifically, the project will investigate and develop novel & scalable networking solutions at three levels. 1. At intra-technology level, IDEAL-IoT will improve the performance of existing LPWAN networks. This objective includes (i) increasing the scalability of existing networks through the design and optimization of advanced PHY and MAC protocols for LPWAN networks; (ii) designing intelligent solutions to support real-time LPWAN traffic with latencies below 100 msec and reliability higher than 99.99%; (iii) improving energy efficiency by a factor 2 through PHY&MAC co-design. 2. At inter-technology level, IDEAL-IoT will improve the performance of coexisting LPWAN networks from different operators as well as provide coexistence between different LPWAN technologies. This objective includes: (i) reducing packet loss due to interference by 50% through interference detection, interference mitigation strategies and inter-technology LPWAN communication, negotiation and optimization; (ii) providing inter-technology roaming and multi-hop communication. 3. At management level, IDEAL-IoT will realize technology agnostic solutions for virtualized LPWAN network management and intelligence. This objective includes: (i) technology-agnostic virtualized components and light-weight APIs for real-time creation of virtualized LPWANs, on-the-fly adjustment of SLAs and dynamic installation of virtualized functionalities to control and improve interactions over LPWAN networks; (ii) the design of a cloud repository capable of optimizing LPWAN settings 10 times faster than is currently the case; (iii) realizing fully reliable over-the-air, reconfigurations and partial software updates of large groups of devices with 50% lower latency and 80% less network overhead.

Researcher(s)

• Promoter: Latré Steven
• Co-promoter: Famaey Jeroen

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The goal of the 5GUARDS project is to investigate, evaluate and demonstrate how various services with diverse requirements can be simultaneously supported by the future 5G network based on the concept of network slicing. 5GUARDS envisions that three building blocks will contribute to the realization of services with diverse requirements: core slicing, RAN slicing and dynamic software reconfiguration.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project website

Project type(s)

• Research Project

Abstract

Experimentally driven research is considered to be a key factor for growing the European Internet industry. In order to enable this type of RTD activities, a number of projects for building a European facility for Future Internet Research and Experimentation (FIRE) have been launched, each project targeting a specific community within the Future Internet ecosystem. Through the federation of these infrastructures, innovative experiments become possible that break the boundaries of these domains. Besides, infrastructure developers can utilize common tools of the federation, allowing them to focus on their core testbed activities.Recent projects have already successfully demonstrated the advantages of federation within a community. The Fed4FIRE+ project intends to implement the next step in these activities by successfully federating across the community borders and offering openness for future extensions.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The Internet of Things is fostering more and more mission critical applications on top of the wireless infrastructure. An example of this is the control of drones, which requires ultra-reliable communication with ultra low latency guarantees and the ability to switch from one technology to the other. Current IoT networks are currently not suited for providing such guarantees as (i) each technology works independently of each other, (ii) applications sometimes have limited control over the devices that are part of the network and (iii) existing high performing management solutions (e.g., Software Defined Networking) only work with resource rich devices. In this project, I propose a way to reach the same level of flexibility in the management of IoT networks as these high performing management solutions offer, without losing the support for resource constrained nodes and third party devices. We do this through WHISPER, an approach that generates "small lies" (fabricated messages) about the network state with the goal of improving the overall network management and providing guarantees on the application delivery. These messages are used to fool the existing IoT MAC, network and transport protocols in such a way that WHISPER will take control over the full network. This includes routing, link and end-to-end communication. As such, WHISPER can be used to manage a multi-technology IoT environment, where mission-critical applications such as drones can be hosted.

Researcher(s)

• Promoter: Latré Steven
• Co-promoter: Famaey Jeroen
• Fellow: Mennes Ruben

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Cities are relying on Internet of Things (IoT) to make their infrastructure smart by using advanced sensing and control devices within the city's infrastructure with the goal of improving urban living, city's experience, etc. Analysis of the data generated by a wide range of sensors and actuators allows controlling the city in a better and more automated way, with respect to e.g. the view on the city's mobility patterns. To realize a smart city infrastructure we consider three layers: the network/sensor layer, i.e. a city-wide network based on a variety of communication technologies and its protocol stacks together with a variety of sensors allowing the collection of raw data; a data layer, dealing with the continuous stream of data and its techniques for processing, storing, mining; an application layer, responsible for interpreting the processed data stream for more optimally controlling the city. The network/sensor layer will be covered by the MOSAIC research group (Dept. Mathematics and Computer Science, Chris Blondia and Steven Latré), while the data layer will be dealt with by the ADREM research group (Dept. Mathematics and Computer Science, Bart Goethals) and finally the application layer is the responsibility of the Transport and Regional Economics research group (Dept. Transport and Regional Economics, Eddy Van de Voorde and Thierry Vaneslander). The general aim of this project is to bring together the expertise present at the University of Antwerp at each of these layers, in order to bundle the research and come up – through an intensive collaboration - with a framework covering the three layers. More specifically, we will build an integrated smart city platform, tailored towards mobility, that allows to capture, process, analyze, interpret and control smart city data in general and mobility data particularly. As discussed in the next section, this will result in important novel research contributions in each of the three layers and will result in a proof-of-concept where the research results are combined into a demonstrator.

Researcher(s)

• Promoter: Blondia Chris
• Co-promoter: Goethals Bart
• Co-promoter: Latré Steven
• Co-promoter: Van de Voorde Eddy
• Co-promoter: Vanelslander Thierry

Research team(s)

• Internet Data Lab (IDLab)

Project website

Project type(s)

• Research Project

Abstract

SynchroniCity represents the first attempt to deliver a Single Digital City Market for Europe by piloting its foundations at scale in 11 reference zones - 8 European cities & 3 more worldwide cities - connecting 35 partners from 11 countries over 4 continents. Building upon a mature European knowledge base derived from initiatives such as OASC, FIWARE, FIRE, EIPSCC, and including partners with leading roles in standardization bodies, e.g. ITU, ETSI, IEEE, OMA, IETF, SynchroniCity will deliver a harmonized ecosystem for IoT-enabled smart city solutions where IoT device manufacturers, system integrators and solution providers can innovate and openly compete. With an already emerging foundation, SynchroniCity will establish a reference architecture for the envisioned IoT-enabled city market place with identified interoperability points and interfaces and data models for different verticals. This will include tools for co-creation & integration of legacy platforms & IoT devices for urban services and enablers for data discovery, access and licensing lowering the barriers for participation on the market. SynchroniCity will pilot these foundations in the reference zones together with a set of citizen-centred services in three highimpact areas, showing the value to cities, businesses and citizens involved, linked directly to the global market. With a running start, SynchroniCity will serve as lighthouse initiative to inspire others to join the established ecosystem and contribute to the emerging market place. SynchroniCity takes an inclusive approach to grow the ecosystem by inviting businesses and cities to join through an open call, allowing them to participate on the pioneering market place enabling a second wave of successful pilots. They will strengthen the ecosystem by creating a positive ripple effect throughout Europe, and globally, to establish a momentum and critical mass for a strong European presence in a global digital single market of IoT-enabled solutions.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The High Impact project of IMEC aims at stimulating fundamental research that can benefit in the long term the valorization of the group. Within this project, the following research lines have been funded: - Appdaptive: configuring IoT networks based on application requirements - Participation in the DARPA Spectrum Collaboration challenge - Densenets: SDN-based network management of wireless networks (resulting in the ORCHESTRA framework) - SubWAN: management of new low power wide area wireless networks

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

In the City of Things initiative, imec, the city of Antwerp and the Flemish Region are working together to make Antwerp a large-scale testbed for the testing and development of smart city technology. With this unique project we want to become a driving force for research into - and the development of - smart cities.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

This project defines the right criteria for an efficient IoT architecture for city logistics. In addition, the added value of the use of this IoT architecture is demonstrated on the basis of specific logistic applications in relation to city logistics.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The immediate cause of the SeRGIo project is the realization by key industrial players that their mobile workforce (e.g. BPost) and customers (e.g. Mobile Vikings/CityLife) are a significant untapped resource that can be used to support new mobile sensing applications. SeRGIo will tap into the huge potential of qualitative mobile sensing applications, and address the complexity (and cost) to develop and deploy mobile sensing applications with state-of-the-art software platforms. The SeRGIo project focuses on qualitative and quantitative mobile sensing at urban scale. This proposed approach complements conventional Smart City platforms with a novel "human as a sensor" paradigm that leverages mobile workforces, groups of incentivised citizens and their mobile devices, to provide sensed data at a far greater scale accuracy and resolution than was previously possible. The SeRGIo consortium argues that to truly unleash the potential of Smart Cities and acquire a complete view on urban dynamics, it is equally important to capture subjective, qualitative metrics in addition to the quantified metrics that are typically collected via physical sensors. Examples of qualitative metrics include perceptions of neighbourhood safety, amiability, cleanliness, subjective experience with urban services, shopping experiences, etc.. The project is motivated by three high-potential business cases (from Nokia, bpost and CityLife), which demand high quality sensed data and professional data analytics. Although the business potential of mobile sensing is clear, it is far from trivial to design and develop a solution that can live up to the expectations of the above business cases. An acceptable solution (1) must collect qualitative as well as quantitative data and exploit synergies by mixing both sources of data, (2) precisely target the allocation of sensing tasks to users based upon their spatio-temporal context, and (3) maximise performance and minimise energy consumption for mobile devices, while ensuring security and privacy. Today's off-the-shelf mobile sensing solutions fail to deliver these requirements. SeRGIo will address them by investigating: 1. A family of domain-specific languages to formalise the business requirements of sensed data customers and inform the orchestration service on how to select the most appropriate subset of mobile users from the participant population. 2. An orchestration service automates the distribution of sensing tasks to a targeted selection of potential end-users based on their location, activities, situational context and device capabilities learned from multi-model behaviour analytics. 3. A flexible and modular data acquisition architecture enables seamless interoperability between platform-independent HTML5 sensing modulesand platform-optimized native accelerators that are downloaded and configured dynamically based upon application demands. This architecture will be populated with a suite of accelerators that use platform-specific features (e.g. Digital Signal Processors, Graphical Processing Units) to achieve orders-of-magnitude improvements in performance and energy efficiency. 4. A modular, lightweight, and multi-model data analytics framework that allows for the on-device assembly of custom data analytics pipelines that combine high-quality quantitative sensor data with insightful qualitative data.

Researcher(s)

• Promoter: Latré Steven
• Co-promoter: Hellinckx Peter

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

This project funds the IDLab valorisation manager (currently Ilse Bracke) to promote the valorisation of research in the IDLab IMEC activities on Internet of Things, 5G and Artificial Intelligence. The primary application domains are smart cities, industry 4.0, mobility and logistics.

Researcher(s)

• Promoter: Latré Steven
• Fellow: Lannoo Bart

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The CONAMO project aims at improving both the training towards and the experience at mass amateur cycling events by continously monitoring and analysing the stream of cycling sensor data generated by the rider and his friends. It does this by introducing both innovations at the network (long-range networks) and the data analysis (machine learning and medical feedback).

Researcher(s)

• Promoter: Latré Steven
• Co-promoter: Famaey Jeroen

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

Wireless sensor networks (WSNs) have become a very popular concept throughout the last decade, due to their wide applicability in monitoring and control applications (e.g. traffic control, environmental monitoring). They are composed of low-cost, battery-powered, constrained and failure-prone sensors and actuators, which are densely, randomly and redundantly deployed, communicating using wireless radio technologies. Recently, the concept of virtualization was proposed for WSNs. It has been applied to facilitate the creation of virtual sensors that provide more meaningful information by combining readings of multiple physical sensors and to support multi-tenancy and sensor hardware reuse by collocating multiple virtual onto a single physical sensor. The goal of this project is to uncover other, unexplored, benefits of WSN virtualization. Concretely, we will develop fully distributed solutions that allow physical sensors and actuators to self-organize into highly resilient and energy-efficient virtual sensing platforms. Resilience will be provided by exploiting redundancy of sensing hardware and network functions within virtual sensors. Energy-efficiency will be improved by intelligently disabling redundant functionality. Optimizing this trade-off dynamically is the main driver of this project proposal. The first two phases will respectively study control aspects within and between virtual sensors. The third phase will extend the solutions to the highly challenging and mostly unexplored field of mobile sensors.

Researcher(s)

• Promoter: Famaey Jeroen
• Co-promoter: Latré Steven
• Fellow: Mennes Ruben

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

As everyday devices are being connected to the Internet, research on large-scale wireless sensors networks specifically and Internet of Things (IoT) generally are becoming more and more important. There is a considerable research and innovation effort related to the deployment of smart cities using this IoT technology. However, there are still plenty of hurdles to move from R&D to implementation and real mass-scale deployment of wireless sensors networks. Moreover, the city itself is a treasure of data to be explored if the right sensors can be installed. Testbeds are the preferred tools for academic and industrial researchers to evaluate their research but a large-scale multi-technology smart city research infrastructure is currently the missing link. The City of Things research infrastructure will build a multi-technology and multi-level testbed in the city of Antwerp. As a result, 100 locations around the city of Antwerp and its harbour will be equipped with gateways supporting multiple wireless IoT protocols. These gateways will connect with hundreds of wireless sensors and actuators, measuring smart city parameters such as traffic flows, noise, air pollution, etc.

Researcher(s)

• Promoter: Blondia Chris
• Co-promoter: Blust Ronny
• Co-promoter: De Backer Charlotte
• Co-promoter: Goethals Bart
• Co-promoter: Hellinckx Peter
• Co-promoter: Latré Steven
• Co-promoter: Poels Karolien
• Co-promoter: Samson Roeland
• Co-promoter: Vandebosch Heidi
• Co-promoter: Vanelslander Thierry
• Co-promoter: Walrave Michel
• Co-promoter: Weyn Maarten

Research team(s)

• Internet Data Lab (IDLab)

Project website

Project type(s)

• Research Project

Abstract

Since the introduction of the first smartphones approximately a decade ago, their popularity has skyrocketed. This leads to an increasing use of IEEE 802.11-based communication, better known as Wi-Fi. As Wi-Fi was originally designed for small-scale home and office environments, it does not scale in face of a growing number of connected devices. There is therefore a need for a large-scale, adaptive Wi-Fi MAC and management framework that can provide the necessary QoS guarantees in the face of highly dynamic environments and is at the same time still compatible with limited scale deployments. Within this research project, we will investigate to what extent we can offer the same QoS in large-scale Wi-Fi networks as can be observed in current limited scale Wi-Fi networks. Large-scale networks in this context can range from hundreds up to ten thousands of connected devices. We will do this by introducing a programmable and adaptive (Software-Defined) way to optimise the MAC layer that support: (i) Application-aware load-balancing for dense Wi-Fi deployments (ii) Adaptive control of MAC parameters to support elastic scaling (iii) QoS estimation in challenged Wi-Fi networks (iv) QoS differentiated reservation

Researcher(s)

• Promoter: Latré Steven
• Fellow: Bosch Patrick

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

AGILE (Adaptive Gateways for dIverse muLtiple Environments) builds a modular hardware and software gateway for the Internet of Things with support for protocol interoperability, device and data management, IoT apps execution, and external Cloud communication, featuring diverse pilot activities, Open Calls & Community building. AGILE builds a modular and adaptive gateway for IoT devices. Modularity at the hardware level provides support for various wireless and wired IoT networking technologies (e.g. KNX, ZWave, ZigBee, Bluetooth Low Energy, etc.) and allows fast prototyping of IoT solutions for various domains (e.g. home automation, environment monitoring, wearables, etc.). At the software level, different components enable new features: data collection and management on the gateway, intuitive interface for device management, visual workflow editor for creating IoT apps with less coding, and an IoT marketplace for installing IoT apps locally. The AGILE software can auto-configure and adapt based on the hardware configuration so that driver installation and configuration is performed automatically. IoT apps are recommended based on hardware setup, reducing the gateway setup and development time significantly.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The High Impact project of IMEC aims at stimulating fundamental research that can benefit in the long term the valorization of the group. Within this project, the following research lines have been funded: - Appdaptive: configuring IoT networks based on application requirements - Participation in the DARPA Spectrum Collaboration challenge - Densenets: SDN-based network management of wireless networks (resulting in the ORCHESTRA framework) - SubWAN: management of new low power wide area wireless networks

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

This project represents a research contract awarded by the University of Antwerp. The supervisor provides the Antwerp University research mentioned in the title of the project under the conditions stipulated by the university.

Researcher(s)

• Promoter: Demeyer Serge
• Co-promoter: Blondia Chris
• Co-promoter: De Meulenaere Paul
• Co-promoter: Hellinckx Peter
• Co-promoter: Latré Steven
• Co-promoter: Peremans Herbert
• Co-promoter: Steckel Jan
• Co-promoter: Steenackers Gunther
• Co-promoter: Vangheluwe Hans
• Co-promoter: Vanlanduit Steve
• Co-promoter: Weyn Maarten
• Fellow: De Mey Fons
• Fellow: Hristoskova Anna

Research team(s)

• Antwerp Systems and software Modelling (AnSyMo)

Project type(s)

• Research Project

Abstract

The central research question of the SMILE-IT project is: "How can complex networks become self- organizing while ensuring stability and without sacrificing on performance. Moreover the decisions taken by the system should be understand- able and guidable." More precisely, the project aims to develop a framework for studying and managing modern distributed networked systems that contain a large number of entities or agents, both machine and human, which strive to achieve their personal objectives. The framework developed within the proposal will guide these entities, either through direct control or by way of incentives, in order to achieve system-wide optimal behaviour, satisfy global objectives and adhere to the system's operational constraints in the face of diverging and incompatible personal goals. Software language abstractions will be identified and developed, to support the ease of the deployment of the framework on a wide variety of networks. The framework will build on the expertise of the teams in machine learning (including game theory, self- organization of complex systems, large-scale multi-agent systems and emergent social behaviour), network management and modelling, and software language design. The key idea of the framework is that the context within which intelligent decision making components or agents operate may depend on spatial and temporal factors. As such, they should be able to adapt their behaviour and goals as a function of space and time. The framework should satisfy the following requirements: It should be generic so as to be applicable to a wide range of networks, it should be scalable with respect to the size of the network, the resulting behaviour should be (near) optimal and at all times minimal performance should be guaranteed, also in unexpected situations. Several fundamental scientific challenges remain to be solved before this high-level objective can be achieved. They can be summarized as follows: - Complex multi-agent control: The SMILE-IT project will develop programming abstractions for distributed network control, that allow agents to be configured and controlled in a network-, rather than agent-centric manner. Moreover, it will provide abstractions to efficiently query and control the state of large-scale complex networks, as reinforcement learning techniques continuously require a view on the current state of the environment. - Fast and stable convergence towards an acceptable solution: SMILE-IT aims to guarantee acceptable performance as soon as the management agents go operational. The project will investigate how to combine learning with heuristic knowledge and existing control strategies, in order to guarantee performance during learning. Moreover, solutions are needed that allow agents with (partially) conflicting goals to collaborate and jointly achieve a converged policy that leads to acceptable performance for all. - Robust and adaptive management under unexpected conditions: Novel learning-based techniques that can cope with large degrees of uncertainty, and select suitable actions even if the network's state is only partially known will be developed. Moreover, unexpected situations, such as failures or faults, may occur during operations. SMILE-IT will investigate how reinforcement learning techniques can be applied to detecting and recovering from such unexpected situations. The development of the SMILE-IT framework will be guided by 2 driving cases in the smart grids and telecom networks application domains. Advanced prototypes for these cases will be developed to support an extensive evaluation of the SMILE-IT technologies. In a later phase, applications to a number of additional domains proposed by the user committee (such as traffic networks) will be examined, in order to demonstrate the general applicability of SMILE-IT methods.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The goal of this research project is to redesign the way services are delivered across the Internet. Instead of focusing on traditional host-based schemes (e.g., TCP/IP-based transport) and best effort delivery models across autonomic network domains, we will investigate how we can exploit both network and cloud virtualization techniques to provide end-to-end QoS guarantees on the service delivery.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The WiSHFUL project (Wireless Software and Hardware platforms for Flexible and Unified radio and network controL) will reduce the threshold for experimentation in view of wireless innovation creation and by increasing the realism of experimentation. More specifically, the WiSHFUL objectives are: To offer open, flexible & adaptive software and hardware platforms for radio control and network protocol development allowing rapid prototyping of innovative end-to-end wireless solutions and systems in different vertical markets (manufacturing, smart cities, home, office, healthcare, transportation, logistics, environmental monitoring...). Key features of such platforms are: Unified radio control allowing full radio control without the need for deep knowledge of the hardware specifics of the radio hardware platform; Unified network control, allowing rapid prototyping and adaptations of network protocol stacks, without the need for deep knowledge of network protocols and software architectures, but also allowing the implementation of novel protocols (e.g. cooperative protocols which require time synchronization and coordination of a subset of nodes); Support for experimentation with intelligent control of radio and network settings, enabling intelligent, node-level and network-wide decisions, on radio and network operation modes and according settings, driven by higher-level domain-specific application demands and taking into account external policies (for example policies for dynamic spectrum access). To offer advanced wireless test facilities that: follow the current de facto standards in FIRE, set by the FED4FIRE project, for testbed interoperability adopting and extending standardized tools for discovery and reservation, experiment control, measurements & monitoring supporting federated identity management and access control; To support diverse wireless (access) technologies and platforms: Create generic and open interfaces for control of the existing devices for technologies like Wi-Fi (IEEE 802.11), Bluetooth (IEEE 802.15.1), WPAN (IEEE 802.15.4), LTE, WiMAX that are already available in current facilities; Extend these interfaces to more open ended experimental radio platforms covering software defined radio platforms, embedded devices and non-commercial grade hardware, so as to enable 5G, Internet of Things (IoT), Machine-to-Machine (M2M), tactile internet; To offer portable facilities that can be deployed at any location allowing validation of innovative wireless solutions in the real world (with realistic channel propagation and interference characteristics) and involving real users. To extend the WiSHFUL facilities with additional facilities or wireless hardware, offering complementary or novel radio hardware/software platforms, supporting experimentation with new technologies such as mmWave (WiGig 60GHz and IEEE802.11ad), full duplex radio, IoT testbeds, smart antennas, etc. To attract and support experimenters for wireless innovation creation targeting different classes of experimenters via different open call mechanisms tailored to the specific classes (industrial relevance for SME versus level of innovation for academia).

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

SHIFT-TV has the ambitious goal of defining the next generation IPTV architecture which completely obsoletes current IPTV systems, while at the same time offering a superior experience compared to OTT (over-the-top) video delivery mechanisms such as adaptive bit rate streaming.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The iMinds IoT program aims at creating a leap forward in Flanders with respect to IoT by consolidating all researchers on one physical location in Ghent. Different research tracks have been defined in the domain of IoT. Within this program, the research group PATS is active in the Cloudlet research line, where we investigate how to develop AI algorithms on top of resource constrained IoT devices.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The goal of the iFEST project is to improve the digital experience at large events such as festivals by developing a new generation of bracelets, which will be integrated with advanced communication and sensor capabilities. Moreover, the necessary festival software platform to both manage the bracelets and analyse the data they generate will be designed. This provides an answer to the "analog way" festival organizers and festivalgoers still experience a festival (i.e., limitedc interaction & communication mechanisms) and will help the market of live entertainment to maintain its strong position within the entertainment sector.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The aim of EMD is to research and demonstrate the core technologies of a media distribution platform that operates across public / private networks with different characteristics and capable of supporting high quality / low latency video collaboration applications in a secure yet user friendly manner.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The Internet has experienced an important evolution since its invention several decades ago. Started as a simple packet forwarder, it is now a delivery platform for rich and demanding services such as cloud applications and video streaming. Despite this important evolution, its underlying architecture has remained the same over the years. Because of this, the Internet lacks flexibility: it is still not possible to provide Quality of Service (QoS) guarantees to Internet services offered by third parties (called Over The Top services) such as Google Apps or Skype. This often leads to severe degradations in the overall quality of the service and corresponding customer satisfaction. A second important evolution is the increasing popularity of cloud environments and their continuous integration in todays Internet. As a result, there is no longer "a cloud" or "an Internet": the two are converged towards one unified platform. In designing the future Internet, it is therefore important to consider the management of a unified cloud and network environment. Because of these evolutions, the research community has been investigating ways of programmatically managing the network (called Software Defined Networking) and virtualizing network resources. Within these activities, the goal is to achieve the same flexibility as can be reached in a cloud environment. Within the area of network virtualization, the concept of Service Function Chains (SFCs) plays an important role. SFCs are graphs, consisting of different sub-services (e.g., a video streamer, a part of a cloud application), which can be distributed and deployed across multiple datacenters and jointly form the Internet service. The construction of these SFCs and the assigning (called embedding) of the SFC components to the different datacenters in a scalable way is an important unsolved problem. The goal of this PhD project is to develop algorithms to construct and embed such SFCs and adapt them dynamically based on user mobility, variable service requirements, and network dynamics. Because of these changes, the optimal SFC construction changes and an on-the-fly provisioning of resources is required. We will consider both optimal algorithms using mathematical optimization techniques and approximations using multi-agent heuristics. By enabling a scalable and dynamic construction and embedding of SFCs, it becomes possible to offer Over The Top services with the necessary QoS guarantees.

Researcher(s)

• Promoter: Latré Steven
• Fellow: Spinnewyn Bart

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project

Abstract

The BeamSat project fits in the development of Newtec's next generation satellite broadband access system using multi-spotbeam Ka-band technology and wideband bent-pipe RF channels. The purpose of BeamSat is to define and develop the next generation broadband Sat3Play solution in order to accelerate Newtec's growth within the broadband market.

Researcher(s)

• Promoter: Latré Steven

Research team(s)

• Internet Data Lab (IDLab)

Project type(s)

• Research Project