Project «PREDIVIS: Hardware accelerated energy disaggregation for energy efficiency and predictive maintenance applications”, funded under the Stavros Niarchos Foundation Programme for Industrial Fellowships. Leader: EC Marcoulaki, PhD candidate: S Kotsilitis (University of the Aegean), Company: Plegma Labs S.A., [2017–2021]

The PREDIVIS project aims to develop novel tools for energy disaggregation and monitoring of device health status. These tools will collect and analyse complex energy load time-series in real-time, using custom edge and cloud algorithms.

Until now, we have successfully developed custom sensors for high frequency sampling, and algorithms for energy disaggregation and fault detection.

The approach has been demonstrated in an industrial environment (see projects SUPREEMO and CLARION) and our systems will soon be deployed at the Nestle premises (new project accepted under the Ignite Ideas II call).

Publications:

1. Kotsilitis S., Marcoulaki E., Kalligeros E., Mousmoulas Y., 2018. Energy efficiency and predictive maintenance applications using smart energy measuring devices. In S. Haugen, A. Barros, C. van Gulijk, T. Kongsvik & J.E. Vinnem (eds.) “Safety and Reliability – Safe Societies in a Changing World”, CRC Press, ISBN 978-0-8153-8682,  pp. 987-994, https://www.taylorfrancis.com/books/9781351174657
2. Kotsilitis S., Marcoulaki E., Kalligeros E. & Mousmoulas Y., 2018. Distributed edge computing paradigm with dedicated devices for energy efficiency and predictive maintenance applications. In “Industrial Internet of Things and Smart Manufacturing”, Springer Series on Lecture Notes on Data Engineering and Communications Technologies (NDECT), in press (ISBN: 978-1-912532-06-3)
3. Kotsilitis S., Marcoulaki E., Kalligeros E., 2019. High Frequency Energy Disaggregation Sampling and Analysis towards Predictive Maintenance Applications. In M. Beer & E. Zio (eds.) “Proceedings of the 29th European Safety and Reliability Conference”, Research Publishing, Singapore, pp. 1214-1222, ISBN: 978-981-11-2724-3; https://doi.org/10.3850/978-981-11-2724-3 0892-cd.