[agents] Post-doc position on reinforcement learning for automated vehicles

[Matthijs Spaan]

Post-doc position (24 months) at Delft University of Technology on model-based
reinforcement learning (RL) in the context of automated vehicles.

The position is embedded in the Horizon 2020 project MEDIATOR 
(http://mediatorproject.eu), which addresses the issue of sharing control
between a human driver and an automated vehicle. The project develops a
mediating system for drivers in semi-automated vehicles, resulting in safe
switching between the human driver and the automated system based on who is
fittest to drive. You will collaborate with experts across Europe to optimize
safety potential of vehicle automation during the transition to full automation.

This Post-doc position focuses on the intelligent decision-making components of
the mediating system. You will be developing novel (model-based) reinforcement
learning techniques in a problem domain that poses strict safety constraints and
involves interaction with both automated systems and human users. You will need
to address key RL challenges such as dealing with several sources of
uncertainty, guaranteeing safe behavior and acting in a manner that is
explainable and intuitive to the user.

You will become part of the Algorithmics research group at Delft University of
Technology. In this group, we aim to design, and understand fundamental
properties of, algorithms for intelligent decision making in real-world
applications. Both within the research group as well as in the broader
university context a significant amount of RL research is taking place,
providing you with ample opportunities for collaboration.

Application deadline: Nov 24, 2019
Further details:
https://www.academictransfer.com/en/287021/

Cheers,
Matthijs