William Clocksin

Autonomous Vehicle

Members of the Intelligent Transport Systems programme are developing technology for autonomous vehicles. Autonomous or “robot” vehicles will be able to navigate and drive themselves. They need to analyse the road automatically to identify road markings, street furniture, and other road users, and to adapt to the road surface. And, they need to be safe around people.
Pasted Graphic

Our test platform is an electric powered autonomous off-road all-terrain vehicle based on a full-size quadbike. The vehicle is intended to be like a “pack mule” that has its own sense of situational awareness around groups of people, and can follow simple commands such as “follow me”, “park over there”, and “unload yourself from the van”. This vehicle is currently under construction in the School of Technology, Oxford Brookes University. A variety of projects is contributing to the design and construction of this vehicle:

  • Visual guidance and navigation. This project is developing real-time visual guidance software and hardware that uses colour stereo image sequences plus information from inertial sensors. People: Prof. William Clocksin and Sunando Sengupta.
  • Intelligent motor control for electric vehicles. The control system needs to use the dynamic response of the vehicle to the motor torque to evaluate the tractive qualities of the terrain it is moving over, to adjust its operation so that traction is maintained, and to ensure that the motors are used at the most efficient torque and speed. People: James Larminie and Marco Cecotti.
  • Road scene understanding. We have access to abundant street level imagery obtained from in-car cameras. The application is to automatically inspect the street level imagery to assay road signs, street furniture, and other roadside assets. Our approach combines appearance information with structure-from-motion features to understand the road scene. People: Paul Sturgess, Karteek Alihari, L’ubor Ladický and Prof. Phil Torr.
  • Heterogeneous wireless broadband access networks.  The communication system uses heterogeneous wireless access technologies. The core component of the test-bed is a Heterogeneous Access Gateway (HAGW). Following technical objectives are addressed: 1) automatic wireless access interface selection; 2) intelligent bandwidth aggregation and allocation; 3) seamless QoS support; and 4) context-aware packet scheduling. People: Dr Shumao Ou and Dr Peter Ball.