Learning Outcomes
The students will have an in-depth knowledge and hands-on experience on emerging fields of Robotics and Autonomous Systems (RAS) towards a smooth transition from theoretical knowledge to their applications in dynamic real-time environments. They will experience a full R&D project cycle covering researching, setting-up, developing and finalizing (testing on real systems or on simulation). These will have a deeper insight into the scientific and technological aspects of existing solutions in artificial intelligence and multi-agent systems. The students will experience various computer vision, learning and planning problems in RAS, will have improved programming skills (e.g. python / C++) in real applications and get familiar with Robotic Operating System (ROS).
Content
The projects comprise of a selection of topics / problems in line with the department’s research and development activities. These changing topics will be from emerging fields of RAS, for example:
- Autonomous Robots in Real-World Tasks (e.g. UAVs, service / companion robots),
- Human-Robot Interaction (HRI),
- Smart Factory (Industry 4.0),
- Autonomous Vehicles,
- …
An introduction to ROS, selected simulation environments (e.g. Gazebo, MORSE) and the available real systems (HWs) will be given in the first few lectures as well as a number of suggested topics. The topics overall cover, but are not limited to, the technical work below:
- complex system architectures / frameworks for autonomous agents,
- multi/single-agent planning (in stochastic/deterministic conditions),
- perception and cognition of sensory inputs (e.g. 3D camera), e.g. object recognition, human & behaviour recognition
- machine learning for real-world applications (e.g. reinforcement learning, deep learning)
- human-aware planning (e.g. assistant robots, autonomous cars)
- basic robotic tasks: motion planning for manipulation
- deployment of a complex real world system
Description of Teaching and Learning Methods
Weekly project meetings, lectures towards the selected topics or used frameworks (ROS), supervised project works in small groups, system design description, milestones, final report and demonstration/presentation.