Cognitive Internet of Things (CIoT)
We work at the intersection between computational intelligence and network science, to address network complexity, high-density telecommunications, and massivescale sensing in the Internet of Things. Complex ICT systems such as Smart Cities will soon generate data o-n the Zettabyte scale. Such data volumes at the network edge will be largely unstructured, uncorrelated, noisy and incomplete, posing Data Science problems that current methods cannot master. We have made breakthroughs in distributed data mining, miniaturized machine learning, and networks mining, with practical applications to Smart Cities, Environmental Monitoring, Smart Energy, Smart Health, and Cyber-Physical Systems.
We have extensive experience in collaborating with the Telecom industry, including projects on: cognitive IoT communications; energy efficiency; software defined networks; user localization; Quality of Experience management; and cognitive video streaming.
The ambition is to connect 1 trillion objects in the Internet of Things. At this scale, connectivity and data mining are distributed prediction problems which start inside the devices. We have recently achieved miniaturized machine learning processes, such as anomaly detection and interference prevention, in sensors having as little as 10 Kbytes of memory.
- Inter-IoT is an EU funded project with focus on open data and interoperability (ranked #1 in the EU Horizon 2020 ICT-30 domain). We have pilots in the Port of Valencia (Smart Logistics) and at Turin Hospital (Smart Health).
- The Joint Intellisensing Lab connects our smart-sensing facilities to an ocean monitoring platform in Shanghai, and an e-health system in Brisbane.
- ACCUS, EU Artemis projectdeveloped a smart city pilot in Gdańsk, Poland.
- DEMANES, EU Artemis project developed a smart logistics pilot in the Port of Rotterdam, the Netherlands.
We maintain an Open IoT facility that allows deployment of applications, generation of datasets, and IoT Big-Data analytics.