Thesis Topic: Development of Interactive Diver-Drone Robotic Features to Assess Divers’ Cognitive Abilities for the Prevention of Diving Accidents

This PhD, funded by the Franco-Italian Maritimo Fabis project, aims to develop technologies to enhance professional diving. It will begin in September or October 2025 and will take place at the COSMER robotics laboratory and the J–AP2S STAPS laboratory at the University of Toulon.

Underwater interventions are often organized by depth range, with divers typically operating within the first 50 meters, and underwater robots deployed at greater depths. However, even within the divers' operating range, drones can provide valuable support. For instance, they can act as tool carriers, offer remote visual perspectives, enhance available information, or provide lighting in hard-to-reach areas. They may also serve safety functions by triggering alerts when necessary.

The gas mixtures used in diving can cause cognitive impairments, which may affect work quality and compromise the safety of the diver and their team. These effects—such as short-term memory loss or slower reflexes—can be noticed by the diver and sometimes by external observers. The objective of this PhD is to continuously assess the diver’s cognitive state through interaction with a companion drone. This drone could detect signs of cognitive decline using an underwater gesture-based communication system, analyzed through deep learning techniques.

One of the main challenges is to detect the diver’s condition without interfering with their equipment or tasks. Another challenge is to develop a quick and simple interaction method between the diver and the drone to confirm cognitive capacity loss. The method will first be tested at various depths in controlled environments, then in real-world conditions at sea. This thesis involves the use of classification methods and gesture recognition, which must be integrated into an autonomous robot operating in an underwater environment.

Candidate Profile

We are looking for a candidate with a strong background in computer science and programming, with in-depth skills in signal processing, computer vision, and data-driven machine learning. Prior experience in robotics (ROS, Linux environments, sensors, control systems) and knowledge of underwater environments will be highly valued.

To apply fill this form