Dr Andreas Hula scientist at Austrian Institute of Technology (AIT) lecture, 6 February 2026
Road users need to be safe at all times during their travels. To achieve this, mobility infrastructure needs to be safe, resilient and well maintained. While there is a wide consensus on the desirability of these goals, the question of how to assess and model safety in particular is a substantial scientific and technological challenge. The lecture will present the usage of the Austrian Institute of Technologies’ (AIT) machine learning- (ML-)based solution “Mobility Observation Box” (MOB) in several use-cases and scenarios of safety model building. The MOB is an easy to deploy image-based solution associated with an ML based pipeline to derive safety indicators and enable mathematical model building, whilst being conformant to European data protection laws (GDPR). The obtained data allows to evaluate a range of different safety topics. In particular, the safety of vulnerable road users is a focal point of MOB usage, be it through analyses of road user flows, distances, surrogate safety measures or before-after analyses of safety measures after infrastructural adaptations. Deploying the MOB is simple and cost-efficient. In this lecture, we will demonstrate how to adjust the MOB-usage and its outputs to tackle the otherwise hard to quantify relevant safety problems of our days.
Speaker Bio:
Dr Andreas Hula studied Mathematics in the Natural Sciences at Vienna University of Technology from 2001 till January 2006. Obtained a PhD in Mathematical Finance, on tractable interest rate models, in January 2010. After research activities in Mathematical Finance (on portfolio selection using control theoretic approaches) at Dublin City University (2010-2012), a position as research associate (post-doctoral researcher) at University College London (2012-2017). From December 2017 has been working as a Scientist at Austrian Institute of Technology (AIT), first in the Center of Mobility Systems (2017-2020), then in the Center for Low-Emission Transport (since 2021) as a member of the Competence Unit for Transportation Infrastructure Technologies in both instances. Main area of expertise and focus of current work is building stochastic models for road safety (risk models) and infrastructure maintenance (predictive maintenance) based on probe vehicle data, aiming to improve safety, comfort and integration in digital representations of road infrastructure.
