Professor of Control Technology
The methods developed by Dimos Dimarogonas enable robots to navigate independently, with the help of built-in computers. This may relate to land-based vehicles, aircraft or submarine vessels, with the capacity to grasp and manipulate items. The research primarily concerns creating mathematical algorithms for this automatic control, and planning tasks that can be performed by autonomous systems with several robots working together. In the future, systems of robots and other machines that can work together, called multi-agent systems, may be of benefit in e.g. office environments and industrial processes, or for deployment in search and rescue operations.
The biggest challenge is how to design appropriate control algorithms for each robot. The aim may be individual for each robot, such as navigation between two points while also avoiding collision with other robots and obstacles. This can also concern grasping and transferring an object of interest, in cooperation with two or several robots. Subsequently each robot is treated as a dynamic system. Here, the aim is to mathematically design control units which respect the robot dynamics, their limited information exchange with the other robots, and uncertainty due to their limited knowledge of the environment in which they operate.