Abstract:

Smart management of energy flows in buildings in combination with renewable energy sources and EV charging can not only reduce energy consumption and CO2 emissions, but also provide a rich set of challenges for Computational Intelligence and data science methods in general.

At our R&D campus a testbed, called Smart Company, for research on energy management, data monitoring and new energy-related businesses was built, extending the facility by Photo-Voltaic (PV) systems, charging stations, 2nd life stationary batteries, H2 electrolyzer, and other elements. More than 100 sensors were installed that measure energy flows in the building. The measurement data was used to identify savings potentials, develop novel anomaly detection methods, and fine tune a digital twin of the building. Parts of this dataset will be made available to the public and I will explain the substantial challenges in checking and cleaning the data.

We have also developed new concepts for energy management, including a smart charging manager, that is now in operation for several years and uses a Model Predictive Control approach.

Many-objective optimization methods were used to identify the optimal sizing of different modules (for example the peak power of the PV system). I will report on several challenges we encountered and new research directions, including the topics of fairness and trust.

Keywords: Energy Management System, Smart Charging, Data Monitoring, Anomaly detection, Model Predictive Control, Business Case analysis

Biography:

Tobias Rodemann studied physics and neuro-informatics at the U Bochum in Germany. In 1997 he did his diploma (master thesis) on evolution strategies. After that he joined Honda R&D in 1998 as a scientist working on brain-inspired computer vision and multi-layer neural networks.

In 2003 he received his PhD from Bielefeld University, Germany, in computational neuroscience. Afterwards his focus shifted to biologically inspired methods for sound localization and audio-visual scene representation.

In the last 10 years his main research interests were in the field of many-objective optimization, multi-criteria decision making, system simulation (digital twins), energy management, and more recently human-machine interaction. In addition to leading a research team he is also managing an internal consultation service for data science. Most of recent research work focuses on topics in the domains of energy and new mobility services.

Tobias is currently a chief scientist at the Honda Research Institute Europe. He contributed to around 100 scientific publications and patents and has an h-Index of 20 according to Google Scholar (in August 2024).