Shock treatment

Scientists at Portsmouth University are deploying their expertise in artificial intelligence (AI) to help Hampshire-based PML Flightlink develop wheels that learn as they are driven.

The 'smart wheels', which are being developed under a £200K DTI-funded knowledge transfer partnership project, will be incorporated into an electric vehicle (EV) created by PML Flightlink.

A modified Mini, it has had components such as transmission and mechanical drivetrain removed to create an EV that incorporates independent quad electric drive and traction control, plus anti-skid built into each wheel. The EV also employs regenerative braking to recover energy.

The vehicle is powered by four separate wheel-mounted motors rather than an engine. Each motor produces 160 brake-horsepower. The combined 640bhp provides acceleration of 0-60mph in 4.5 seconds and a top speed of 150mph.

A small, 250cc petrol engine charges the battery while the car is being driven. In this mode, it will run for up to 900 miles before needing to re-fuel, while in pure EV mode it will run for 200 miles. Previous electric models barely managed 60mph and had a range of less than 100 miles.

With a performance of 80-100 miles a gallon compared with 40mpg for the average car, such an electric vehicle will be cheap to run and, with few mechanical parts, cheaper to maintain.

The motors can run as part of a fully electric vehicle or a hybrid system, with a combustion engine powering a generator that charges the vehicle's battery.

'The idea was kicking around for a number of years,' said Martin Boughtwood, owner and director of the company.

'We had the belief that we could create a very efficient electric vehicle but could see no one was approaching it in the right way. There is a stigma attached to electric vehicles but really they can outperform combustion vehicles.'

Though the original system already featured anti-slip braking and traction control, PML Flightlink is now working with the University of Portsmouth to add artificial intelligence, creating a smart wheel that can sense road conditions and modify the car's performance to compensate for problems.

The completed vehicle will have no transmission as all its controls, including suspension and braking, will be linked using software. However, this will not be a drive-by-wire vehicle — the driver will remain in complete control at all times, allowing him or her to still experience the pleasure of driving.

The wheels will use micro- computers to perform 4,000 calculations a second and communicate with each other. The AI will enable the system to learn as the car is driven, making calculations and adjustments according to travelling speed and road conditions. This will give tighter control, a smoother ride and a safer drive, though the driver will remain fully in control of the car.

The AI will manage the suspension, steering and braking systems, teaching the vehicle to adapt to bends in the road, potholes and other potential hazards, compensating by adjusting the car's reactions.

The system will retain information in its memory and use it the next time the car encounters similar road conditions. This will allow it to respond faster and improve the ride for the driver and passengers. The vehicle therefore learns as it drives and adapts its performance to suit what it knows about the road.

'If the car experiences familiar conditions it can compensate by controlling the linear motors in the suspension,' said Dr David Brown of the University of Portsmouth's Institute of Industrial Research. 'It can also predict problems. If it encounters certain conditions it can predict that it is on a slippery road and prevent skidding.'

Electronic traction control and suspension will counterbalance the drop and bumpy ride of a poor road surface, while the drag experienced on a tight corner will also be eliminated, as the car will tilt into the bend. The vehicle will therefore also be safer.