Paul Havinga

Professorship chair

Pervasive computing systems and their applications (University of Twente).

Pervasive Systems are systems composed of a network of collaborative sensing, computational, and reasoning components that are highly embedded in and actively – yet unobtrusively – interact with the environment.

Embedding millions of networked smart sensing objects into an environment creates a digital skin which senses its immediate space. Such distributed networks of smart objects cooperate to support an application as unobtrusively as possible (transparency), making efficient use of scarce resources independent of growth (scalability), in such a way that the system adapts to a dynamically changing environment (evolvability), and that operates and gives results that can be relied upon (trust). Mainly due to resource constraints, devices and connections are inherently unreliable, yet the system should be able to provide reliable services (quality). The context of this research is smart sensor systems, Internet of Things (IoT), cyber physical systems, wireless sensor networks, and smart objects (e.g. wearable, smart phones).

We perform research in pervasive computing systems and develop groundbreaking innovative and smart sensing solutions in application domains such as healthcare, mobility, logistics, industry and manufacturing, biodiversity and energy. The combination of our leadership in smart sensing systems, wireless networking technologies, and embedded AI expertise is what makes us unique. We address the following research themes:

• Smart sensing systems: seamless integration of sensors, computation, and communication, IoT integration, wireless sensor networks, and innovative signal processing with machine learning algorithms to improve accuracy, robustness, and energy efficiency.
• Wireless networking technologies: wireless networking in challenging environments such as underwater, underground, intra-body, in flight and in space, opportunistic networks, urban/mobile crowd sensing & intelligence.
• Embedded AI: models and algorithms, deep learning, activity recognition, context modeling and reasoning, applied machine learning, convergence of IoT and Big Data.

Research questions cover architectures, protocols, programming paradigms, algorithms, and applications. Our system-oriented research is inspired by real problems, which makes multi-disciplinary collaboration natural. In this context, my research contributes closely to the TNO roadmaps on ‘Fast and open infrastructures’, ‘Data sharing, ‘Trusted ICT’ and ‘Embedded Systems Innovation’.

Top publications

• Wei Wang, Fatjon Seraj, Nirvana Meratnia, Paul JM Havinga, “Speaker counting model based on transfer learning from SincNet bottleneck layer”, 2020 IEEE International Conference on Pervasive Computing and Communications (PerCom)
• Le, D. V. , Meratnia, N. , & Havinga, P. J. M. “Unsupervised Deep Feature Learning to Reduce the Collection of Fingerprints for Indoor Localization Using Deep Belief Networks”. In IPIN 2018 - 9th International Conference on Indoor Positioning and Indoor Navigation (pp. 1-7).
• Bosch, S., Serra Bragança, F. , Marin-Perianu, M. , Marin-Perianu, R. , van der Zwaag, B. J., Voskamp, J. , Havinga, P. “Equimoves: A wireless networked inertial measurement system for objective examination of horse gait”. Sensors (Switserland), 18(3), 2018.
• Seraj, F. , Meratnia, N. , & Havinga, P. J. M. “RoVi: Continuous transport infrastructure monitoring framework for preventive maintenance”. In 2017 IEEE International Conference on Pervasive Computing and Communications, PerCom 2017 (pp. 217-226).
• M Shoaib, S Bosch, O Incel, H Scholten, P Havinga, “Complex human activity recognition using smartphone and wrist-worn motion sensors”, Sensors 16 (4), 2017
• A Mijuskovic, A Chiumento, R Bemthuis, A Aldea, P Havinga, “Resource management techniques for cloud/fog and edge computing: An evaluation framework and classification”, Sensors 2021