ICAI: The Labs – AI & Mobility in NL
This month, ICAI: The Labs is focused on AI and Mobility in the Netherlands. The EAISI Mobility lab and the Utrecht AI & Mobility Lab each present their work and discuss challenges and developments made in this field.
The EAISI Mobility lab is a collaboration between Eindhoven University of Technology and NXP Semiconductors. The aim of the lab is to use AI technology in vehicles and transport services to bring about accident-free mobility.
The Utrecht AI & Mobility lab is a collaboration between Utrecht University, ProRail, Nederlandse Spoorwegen (NS) and Qbuzz. The lab focuses on innovative AI techniques to help resolve the challenges facing mobility and public transport.
Program
12.00 (noon): Opening by host Martin Oswald, lab manager Atlas lab (UvA)
12:05: Introduction EAISI Mobiliy Lab by Emilia Silvas (TU/e)
12:10: Manuel Muñoz Sánchez (TU/e) presents “AI in Autonomous Vehicles – A Deep Dive into Motion Prediction”
12.25: Introduction AI & Mobility Lab by Marjan van den Akker (UU)
12:30: Jan de Mooij (UU) presents ‘Studying Effective Behavioral Interventions through Changes in Mobility in Agent-Based Simulations’
12.45: Discussion what’s next for AI & Mobility in NL
13.00: End
Abstracts
“Studying Effective Behavioral Interventions through Changes in Mobility in Agent-Based Simulations” by Jan de Mooij (UU)
In an epidemic, behavioral interventions are often the first line of defense due to the absence of pharmaceutical interventions. At the onset of COVID-19 governments worldwide adopted such strategies of behavioral interventions.
However, such interventions are only effective in so far as they are being adopted by the general population. Due to their design, an important indicator of adoption of these interventions are the changes in mobility patterns.
We have developed an agent-based data-driven model, instantiated from a detailed synthetic population and weekly activity schedules from Virginia, and calibrated on the changing mobility patterns and spread of the disease to study the effectiveness of behavioral interventions. We show through several experiments how changes in intervention strategies helped mitigating the spread of the disease and propose a new method for finding optimal interventions.
Abstract
“AI in Autonomous Vehicles – A Deep Dive into Motion Prediction” by Manuel Muñoz Sánchez (TU/e)
Autonomous Vehicles (AVs) have become widely popular in recent years since they have the potential to increase road safety and bring several additional benefits such as higher traffic throughput, fewer CO2 emissions and improved fuel efficiency. In order to achieve higher levels of automation, an AV must be capable of predicting the actions of other road users and accurately anticipate how the driving environment will evolve.
During this talk we will dive deep into motion prediction for AVs, covering the strengths and weaknesses of various existing methods, identifying current challenges in this line of research, and highlighting interesting research directions to accelerate progress in the field of motion prediction.
