PhD Research Project: Cooperative Behaviour of Autonomous Unmanned Vehicles

When a group of autonomous vehicles are on a mission to explore an unknown area and they cannot be remotely controlled, they need to collaborate with each other on gathering the maximum amount of information. They also need to make sure that they are in contact with each other as a team so that no-one becomes detached from the group. They need to agree about movement plans in realtime. To search for or to examine a finding more closely, they may decide to form a subgroup while others continue what they have been doing. All these actions need to carried out under environmental constraints and to avoid dangers.

The methodology to be used to create such vehicle control software is based on agent oriented programming with policy value functions that the agents optimize in a distributed fashion . The integration of distributed constraint optimization (DCOP) and agent movement and sensing capabilites with realtime evaluation measurements is a major challenge that forms the basis of this research. The project is to use an advanced intelligent agent programming system using natural language programming (sEnglish) with logic based reasoning and planning capabilities.

Prospective applicants should have a computer science or mathematics or control engineering degree and preferably some familiarity with hybrid system modelling and programming in Java/C++/MATLAB.

Funding Notes

Applicants can apply for a Scholarship from the University of Sheffield but should note that competition for these Scholarships is highly competitive. http://www.sheffield.ac.uk/acse/research-degrees/scholarships