Research project title
Safe and Efficient Robot Path Planning Using Formal Conformance Testing for Navigation in Dynamic Indoor Environments
Education level
Doctorate
Director/co-director
Director: Pierre-Yves Lajoie
Co-director(s): Lina Marsso
End of display
August 29, 2025
Areas of expertise
Intelligent systems applications
Modelling and simulation studies
Primary sphere of excellence in research
Modeling and Artificial Intelligence
Secondary sphere(s) of excellence in research
New Frontiers in Information and Communication Technologies
Unit(s) and department(s)
Department of Computer Engineering and Software Engineering
Conditions
-
Familiarity with autonomous systems and robotics
-
Programming experience (e.g., Python, C++)
-
Experience with ML frameworks like PyTorch.
-
Interest in formal methods or model-based testing is a plus
Detailed description
This project aims to develop a novel framework for safe and efficient robot navigation in highly dynamic indoor environments like hospitals, warehouses, and airports. Our approach involves constructing rich, 3D scene graphs that capture the complex spatial, semantic and dynamic relationships within these spaces. The created probabilistic spatial representations will encode the movement of objects, the presence of mobile agents, in addition to time-dependent safety constraints. We will leverage formal methods, specifically model checking, to systematically generate an exhaustive dataset of provably safe and optimal navigation plans. This dataset of "correct-by-construction" plans will then serve as the foundation for training a robust, generalizable deep learning policy, enabling robots to anticipate and react to environmental changes in real-time, ensuring both safety and task efficiency.
Financing possibility
Funding available.
