DC1 - Progressive Evaluation of Shared Models for Human-Robot Understanding

Objectives

1. To devise a methodology for a robot to build and maintain a dynamic model of agents in its social-cognitive environment.
2. To demonstrate its utility through Theory of Mind tasks (e.g., false belief scenarios).
3. To establish operational metrics to assess alignment between the robot’s model and the user’s actual mental state during natural interactions.

Expected Results

1. Symbolic (ontological) models of the environment, validated against ground-truth models through synthetic or manual approaches.
2. Metrics to evaluate the psycho-social alignment between human and robot models over time; add demonstrations of these metrics in actual, real-world scenarios.

Planned Secondment(s)

S1: UBremen Prof. Hochgeschwender, 2 months in M14-15
S2: TUGraz Prof. Steinbauer-Wagner, 2 months in M26–27

Required Skills

Essential

• Degree in Computer Science or related field
• Solid programming skills (Python is essential; C++ is desirable)

Desirable

• Experience in knowledge representation and symbolic reasoning
• Experience in social psychology
• Experience with ROS and simulation techniques
• Practical experience with Machine Learning algorithms