DC13 - Multi-modal Uncertainty Quantification and Management of AI

Objectives

1. To develop techniques for effective and reliable uncertainty quantification of AI components used in robotic applications.
2. To investigate approaches for managing the uncertainty associated with AI components, thus enabling effective and efficient risk-informed decision making.

Expected Results

1. A toolset of techniques for uncertainty analysis and quantification based on the operating context and data.
2. Techniques for effective and efficient uncertainty-based decision making that enable reasoning about epistemic and/or aleatoric uncertainty.

Planned Secondment(s)

1. S25 ITU, Prof. Wasowski, 2 months in M14-15
2. S26 PAL, Dr. Lemaign, 2 months in M38-39

Required Skills

Essential

• Computer Science or Applied Mathematics degree
• Excellent programming skills (C++, Python etc.)

Desirable

• Knowledge in ROS and robot simulators (e.g. Gazebo, IsaacSim, etc.)
• Practical experience with Deep Learning frameworks (Pytorch, Tensorflow) and/or LLMs