Sridath Tula's thesis defense

This thesis presents the design, formalization and feasibility evaluation of an intelligent assistance system (IAS) intended to facilitate human-robot cooperation (HRC). The research introduces a structured framework for designing a cooperation assistant for mixed human-machine teams, combining theoretical modeling, system implementation and empirical validation.

This framework is based on two complementary models: Know-How-to-Operate (KHO), which defines agents' operational skills, and Know-How-to-Cooperate (KHC), which formalizes cooperative behaviors such as information sharing, interference management and task allocation. These models are unified in a grid-based coordination structure that formally represents the interaction dynamics between human operator, IAS and robotic agents, enabling explicable and adaptive decision-making at different levels of autonomy.

To make this framework operational, four ad hoc levels of cooperation are defined: teleoperation, decision support, autonomous decision-making and supervised autonomy, which regulate the balance between human supervision and system autonomy. The IAS architecture was implemented using ROS 2, STRIPS-based planning and a graphical user interface (GUI) to support real-time monitoring and adaptive plan execution in a pick-and-place task inspired by search and rescue operations involving two RoboMaster EP robots. Experimental evaluation, conducted in seven scenarios with 11 participants, confirms the technical feasibility of the framework.

Reporters

Examiners

• Dimitri MASSON. Assistant Professor. ESTIA-Research, ESTIA.

Thesis director

• Marie-Pierre PACAUX-LEMOINE. HDR Research Engineer, LAMIH UMR CNRS 8201, UPHF.

Thesis co-supervisor

• Emmanuelle GRISLIN-LE STRUGEON. Professor. LAMIH - UMR CNRS 8201, INSA Hauts-de-France.

Human-robot cooperation; intelligent assistance system; human-robot team; STRIPS planning; search and rescue; levels of cooperation.