Keynotes
During this year's EMISA, there will be two keynotes related to the discussions aimed at the potential joint bi-lateral and/or EU projects in the area of future mobility.
Keynote 1: Energy challenges in future transportation
Speaker: Dr. Marcin Seredynski. Head of innovation and research, E-Bus Competence Center (Livange, Luxembourg)
Abstract: The transition to electromobility is argued to reduce energy usage, as well as to cut emissions of greenhouse gases, pollutants, and noise. However, the shift to electric vehicles is facing several challenges, especially related to electricity supply. This presentation will first present the status of public transport electrification and possible scenarios for the future. Special focus will be made on the quantification of environmental and societal benefits related to different driveline technologies available today on the market. In its second part the presentation will focus on today’s deployment dilemmas and barriers that are limiting the pace of the transition to electromobility. Energy-related challenges will be discussed by using the example of fuel cell and battery electric vehicles.
Speaker bio: Dr. Marcin Seredynski is head of Innovation and Research at the E-Bus Competence Center in Luxembourg, Marcin Seredynski is involved in R&D projects in the field of cooperative intelligent transportation systems. He is also working on several deployment projects of electrified buses in Europe, Asia, and Australia. Mr. Seredynski received his M.Sc. (2004) from the Warsaw University of Technology and in 2009 he defended his PhD thesis with the University of Luxembourg and Polish Academy of Sciences. He has published over 60 peer-reviewed scientific and technical articles. Mr. Seredynski is honorary chair of UITP.
Keynote 2: Assessing ADAS/AD system's real-life performance: Uniting real data, AI and optimization for scenario-driven design validation and optimization
Speaker: Alex Mugnai, ESTECO Automative
Abstract: Autonomous vehicles should safely manage every possible situation on the road. The proposition is to address vehicle validation with data-driven scenario-based development. Naturalistic driving data and statistical tools allow extracting a comprehensive scenario catalogue, including edge cases, that is in turn used for AD feature design validation and tuning. Data collection in test vehicles is expensive, therefore we present a faster, more affordable, and free of biases solution for scenario data collection.
A demonstration on the method’s application is performed, where we use data from infrastructure sensors to build behavior models (exploiting inverse reinforcement learning algorithms) which we integrated with an open-source simulation environment. We demonstrate how large amounts of synthetic scenarios are generated and how the statistical similarity to observed interactions is evaluated.
The approach, therefore, proposes generation of synthetic driving scenarios with realistic, non-deterministic behavior models for design validation and optimization of a given ADAS/AD function
Speaker bio: Alex has studied biomedical engineering at the Politecnico di Torino in Italy. After a number of year working on vehicle passive safety systems in the USA and Europe, Alex has been involved in the ADAS/AD domain for the last 15 years. He is today responsible for business development at ESTECO where the provided technologies are proposed to support the validation of ADAS/AD functions in collaboration with a number of other companies he is working with.
