Abstract
Automated Driving Systems (ADS) will entail major societal changes. I shall explore some relevant issues of interest to the OR/ML/Stats communities and provide some solutions derived during a recent EU project. Specifically, I shall discuss four topics: The request to intervene decision; providing ethical decision support; the problems with heterogeneous fletes; robustifying ADS algorithms against adversarial attacks. For each of them, I shall specify the problem, provide a novel solution approach and discuss numerical solutions within simulation environments.
Bio
David Rios Insua is research professor in Risk Analysis and Data Science at the Institute of Math Sciences (National Research Council, Spain), Member of the Spanish Royal Academy of Sciences and Professor of Statistics and Operations Research at Madrid Complutense University (on leave). Previously he has taught and/or researched at Purdue University, Duke University, SAMSI, Aalto University, Paris Dauphine, Manchester, Leeds, IIASA and USST. He is DeGroot Prize from ISBA, Edelman Laureate from INFORMS and AXA-ICMAT Chair and received best paper awards from the Risk Analysis, Decision Analysis Societies and the Spanish OR and Stats societies. He has supervised 27 PhD students and managed over 17M€ in research funds and has collaborated in setting up several startups. His interests are in Bayesian statistics, Decision Analysis, Risk Analysis and Machine Learning and their applications in Security, Cybersecurity and Health. He is an avid surfer in the summer.