University’s expertise advises on West Cambridge Site trial of self-driving shuttle

The University’s expertise in sustainable transport technology is supporting a West Cambridge Site trial of a fully electric Autonomous Vehicle.

Suitably sci-fi in appearance - prompting double takes from passers-by and offering a striking vision of the future - the 12-seater self-driving shuttle has been developed by engineering firm Aurrigo and Smart Cambridge, a workstream of the Greater Cambridge Partnership (GCP), under the guidance of the Department of Engineering’s Professor John Miles.

“It wasn’t so long ago that if you said ‘electric vehicles’, the first thing people thought of was a milk float,” said Professor Miles, Director of Research in Transitional Energy Strategies. “We didn’t want it to look like a milk float! 

“Autonomous Vehicles (AV) are about encouraging people to get out of their cars and use public transport. Apart from the technical requirements you want to present something that’s comfortable and hopefully inspiring. Something that doesn’t make you feel like you’ve gone to a lot of trouble to leave your car behind - we want to break the mould of people’s thoughts about public transport.”   

Professor Miles, whose work looks at transport energy, future transport systems, and pathways to a zero carbon future, helped bring the project partners together for the trial, which is investigating the technology’s potential to offer out-of-hours travel from city centres and employment sites as part of a public transport service.

“This is about exploring how an AV system might work in a city like Cambridge. The big opportunity that AVs offer is that they don’t need drivers, so we can have more vehicles running more frequently. It means people don’t have to wait at the bus stop, and they don’t need to worry about a timetable.

“It offers an attractive service which means you don’t mind leaving your car in the Park and Ride, because you have something equally good that is part of a shared transport system.” 

By continuously processing its surroundings as it moves along the road, the shuttle is able to cross-reference fixed objects, such as kerbs and signs, with a map uploaded to its on-board computer. When it detects moving objects on its trajectory it can respond appropriately.

“In very simple terms, the vehicle makes decisions akin to what a human might do,” said Professor Miles. “It’s scanning for things that might intercept its path, and at each stage it’s calculating, in fractions of a second, whether it needs to adjust its speed, whether it needs to stop, whether it needs to turn to the left or right.”

And it is the algorithms behind these decisions, and making a computer powerful enough to run them and small enough to package into a shuttle, which is key to taking these vehicles from the test track to the open road.  

“Right now in the AV industry there are very bright computer programmers putting these algorithms together, trying to outperform one another and produce a system that balances capability and size. We’re very close to that point, but of course the technology needs to be developed in sync with regulators and legislation to create a real-world framework. Once that happens, once that tension between capability, price and size is solved, the technology will take off as a commercial offering.”