UK's RobotCar demonstrated (w/ Video)

Feb 15, 2013 by Lin Edwards report
UK's RobotCar demonstrated

(Phys.org)—A group of scientists at Oxford University in the UK have demonstrated their version of the self-driving car, which promises to be much cheaper than similar cars being developed elsewhere.

The car, being developed by Oxford RobotCar UK, is designed to allow drivers to decide to let the car take over the driving when the car's computer makes the offer. A simple tap on the brakes at any time returns control immediately to the driver.

Unlike other autonomous vehicles being developed by groups such as , the UK team have concentrated on developing their technology using cheap components such as sensors to dramatically reduce the cost. At present, the system costs about £5000 (7750 USD), but the scientists hope to be able to reduce the cost to as little as £100 (about 150 USD)

This video is not supported by your browser at this time.

The system consists of a modified Nissan LEAF electric car with an iPad mounted in easy reach of the driver, lasers attached to the front of the car, and stereo cameras that work with the lasers to produce a three-dimensional image of the route being taken. The lasers and cameras constantly map out the route being taken by the car.

The entire system is controlled by a Main Vehicle Computer (MVC) installed in the trunk, which can control virtually every aspect of the car. A third computer is the Low Level Controller (LLC), which communicates directly with the iPad user interface. The system does not depend on GPS.

This video is not supported by your browser at this time.

The goal of the project is for the RobotCar to drive itself mainly on short journeys on routes with which it is familiar. During the initial period the RobotCar builds a 3D map of routes taken by the car, which is being driven normally. Once it has built up a memory bank of familiar routes it begins to offer to take over and the driver can accept by tapping on the iPad screen. If there is any conflict between the three on-board computers autonomous driving is not offered. If there are any problems during autonomous driving the prompts the driver to resume control. If they do not, the car slows to a stop.

A under the front fender scans with an 85 degree field of view up to 50 metres ahead for obstacles such as pedestrians, other vehicles, or objects, and if the is in control it automatically slows to a stop if an obstacle is detected. It accelerates and continues its journey only when the road ahead is clear.

The Oxford team, led by Professor Paul Newman, expects manufacturers will be fitting the system into their cars within 15 years or less. They say the will make the roads safer and less congested, and will be particularly useful for drivers following regular routes. Professor Newman said the cars of the future would be driving themselves some of the time rather than all of the time.

This video is not supported by your browser at this time.

The RobotCar has been demonstrated on private roads around the Begbroke Science Park in Oxfordshire, but the group is in talks with Britain's Department of Transport to arrange for trials on public roads.

Explore further: Intelligent materials that work in space

More information: mrg.robots.ox.ac.uk/robotcar/

Related Stories

Robot car to cut jams & prangs

Oct 10, 2011

Robotic car technology being developed at Oxford University that interprets its surroundings and makes decisions about where to go could eliminate the agony and cost of traffic jams.

Toyota, Audi driverless demos will pull up to CES

Jan 06, 2013

(Phys.org)—While the Consumer Electronics Show in Las Vegas starting January 8 will be full of mobile-computing gadgetry next week, autonomous driving demonstrations will also capture visitors' attention, ...

Recommended for you

Intelligent materials that work in space

Oct 23, 2014

ARQUIMEA, a company that began in the Business Incubator in the Science Park of the Universidad Carlos III de Madrid, will be testing technology it has developed in the International Space Station. The technology ...

Using sound to picture the world in a new way

Oct 22, 2014

Have you ever thought about using acoustics to collect data? The EAR-IT project has explored this possibility with various pioneering applications that impact on our daily lives. Monitoring traffic density ...

User comments : 5

Adjust slider to filter visible comments by rank

Display comments: newest first

Aryeh_Z
1 / 5 (1) Feb 15, 2013
How does it know to stop at traffic signs or lights?
dschlink
4 / 5 (2) Feb 15, 2013
Traffic signals and signs are rather easy to identify. They are uniform in appearance, placement, and there are a limited number of options.
Eikka
not rated yet Feb 15, 2013
Professor Newman said the cars of the future would be driving themselves some of the time rather than all of the time.


This is a better alternative than making totally self-driving cars, because then people won't forget how to drive when the computers fail them.

Eikka
5 / 5 (1) Feb 15, 2013
Traffic signals and signs are rather easy to identify.


For humans.

For computer vision, we're still at a phase where painting a black circle on your cheek confuses face detection algorithms, and road planning is sometimes so poor that even we have troubles interpreting which of the lights applies to us, or what part of the road the sign is for.

Maybe if there was a way to teach the computer how to drive, by pointing out on the iPad screen "This is the traffic light here; wait till it's green. Watch out for crossing traffic here. The speed limit is 60 here."
Eikka
5 / 5 (2) Feb 15, 2013
One tricky situation is when you're at an intersection, and you're on the direction that has the right of way, but you have no idea that you have.

You have to watch for the backsides of the yield signs on the crossing roads to determine that they have to yield and not you.