Miniature car that could change the future of driverless vehicles

Jon Lawson

Final year engineering students, supervised by Dr Alistair McEwan, at the University of Leicester, have developed a miniature autonomous drive-by-wire vehicle that will provide knowledge that could speed up the development of driverless cars.

The student-developed ‘μPod’ prototype could potentially lead to a whole new system of testing code and algorithms for connected autonomous vehicles (CAVs).

Fully autonomous vehicle are ones in which a driver is not necessary, although these vehicles will be able to carry passengers. Connected vehicle technologies allow vehicles to ‘talk’ to each other and to the infrastructure around them, with well-known examples including GPS navigation.

The vehicle prototype, developed by five final year Engineering students at the University of Leicester, is a 1:6 scale version of the LUTZ Pathfinder Pod developed by Transport Systems Catapult (TSC). TSC is a not-for-profit Technology and Innovation Company, which undertakes applied research projects in collaboration with academia, SMEs and industry with the aim of making the UK a world leader in transport innovation.

The students’ drive-by-wire miniature vehicle has been created to act as a representative initial platform for the testing of the code and algorithms, which would then be applied to the full-size pod. It is capable of driving under remote control, but also has the electronic components required to enable autonomous control.
 
Remote and autonomous control

The vehicle can function in two modes: by remote control using a standard Xbox controller; and under autonomous control. The autonomous mode enables the vehicle to follow a previously mapped out route without further command input.

This works by first recording a GPS route for a certain route through the Xbox remote control. After a completed GPS route recording is collected, the autonomous mode can work by following the recorded GPS file with latitude, longitude and heading as reference, allowing it to move to the destination with self-adjustment as necessary.

The five MEng students behind the prototype are Daniel Thomson, Jingyu Chen, Scott Baker, Mohamed Khaled and Hongfei Chen.

Daniel Thomson said: “The benefit of testing using a prototype such as ours is that it significantly reduces the risk of testing new code, in particular the obstacle detection and avoidance functions, as the overall cost to fix damage or reproduce our vehicle would be a lot lower than, for example, the LUTZ Pathfinder.

“We hope that the vehicle can be used as a platform to test new algorithms and coding for the purpose of enabling autonomous control of a full-size vehicle capable of carrying passengers.

“The completion of this project will allow research to continue in this sector both within the University of Leicester and through our connected company TSC. As interest and awareness of the possibilities involved with autonomous vehicles (improved safety, better congestion control) grows, it will allow advancement to move at a faster rate and make widespread autonomous vehicles a reality sooner than many would think.”

He adds: “Also, if the use of a miniature version of the full size vehicle is successful for TSC, then other companies in the autonomous sector may also look into the idea of using a similar method to test their code, potentially resulting in a whole new system of testing started by the University of Leicester.”

Project supervisor, Dr Alistair McEwan, said: “Connected and autonomous vehicles (CAVs) are appearing on our roads and in our society – they are no longer a thing of the future: they are very much a thing of the present. UK government is aiming for widescale deployment on our roads by 2021, with the ambition of putting the UK at the heart of the international transport innovation sector. Safety, and in particular reasoned arguments about safety, is integral to this.”

He added: “At Leicester we have renowned expertise in validating safety arguments about software and machine learning in safety-critical systems – and projects such as μPod allow us to study new and innovative methods of verifying the safety of these vehicles as their uptake becomes widespread.”

Recent Issues