Last updated 11th August 2004 ACW
Development of sensory and control algorithms to control an autonomous public-transport vehicle in unsegregated environments
This project is to develop an autonomous control system for a public transport vehicle that operates in an unsegregated (or partially segregated) environment. Whereas fully automatic vehicles already operate on totally segregated guided rights-of-way (Metros, "people-movers", lifts) and the major problems associated with driverless vehicles have been solved, research has concluded that fully automatic road vehicles are not viable at this time. This is mainly for non-technical reasons (public acceptance, safety compliance, legal issues). To limit these problems, it is proposed to develop a sensor and control system primarily for a driverless light rail vehicle (tram).
As well as using the proven advantages of this mode of public transport (accessibility, adaptability, comfort, environmental-friendliness), the fixed path involved minimises the complexity of the required algorithms and also assures other road users of the corridor where the autonomous vehicle operates. It is envisaged that the system will primarily be used in combination with inexpensive ultra-light rail technology (shallow rail technology, standard vehicle components) to produce a low-cost flexible system that can be adapted in any environment. The sensor and control systems developed will also have applications in other situations (safety aids for vehicles with human drivers, robot vehicles in large warehouses etc.)
Michael Bell (Safety Consultant)
Lewis Lesley (Prof. of Transport Science, Liverpool John Moores University)
Adam Winstanley (academic, CS, NUIM)
TRAM Power Ltd, Liverpool (tramcar manufacturer)
Model test vehicle
Sensors and signal processing
Computer vision for obstacle detection
Fusion of sensory data
Formal/Rigorous safety-critical software development
Resources
The possibility of a robot street tram (Michael Bell)
Sensors on trams for collision avoidance abstract (Markham, McLoughlin, Winstanley), to be presented at Light Rail 2004
Using a LIDAR scanner to highlight obstacles of interest in image data Tech Report (McLoughlin)
Tram Power Ltd., Tram Power City Class Trams, 2003, http://www.trampower.co.uk/