What is the ACTIVE project?

The ACTIVE project (Advanced Camera Technology in Visual Ergonomics) will focus upon the operation of secondary in-vehicle controls in cars (e.g. radio and climate controls). The project is funded by the EPSRC and the DTI under the Foresight Vehicle LINK programme. It is being undertaken in the Department of Applied Computing at the University of Dundee, in collaboration with other project partners.

Analysis has shown that a proportion of motor vehicle accidents are caused by driver inattention that is induced by the operation of secondary controls. The project will develop a novel, finger-pointing interface for operating such controls. The current physical secondary controls will be replaced by a computer interface, operated by the driver using pointing gestures that are identified by a computer vision system. This approach will utilise digital cameras and image processing. A number of potential safety benefits may accrue from this new interface, for example:

• Faster interaction with secondary controls;
• The ability to situate the computer display in a position much closer to the driver's normal line of sight;
• The interface configuration can be taken from vehicle to vehicle (avoiding the potential safety problems that arise during the period where a driver adapts to the layout of controls in a new vehicle).
• The interface can be operated with both hands still on the wheel.

Our new system would also avoid the complex wiring looms required for current secondary controls, which have high associated labour costs.

The prototype interface currently uses an electromagnetic sensor to determine the direction of the pointing gesture. This will be replaced by a computer vision system which is being developed in parallel with the interface. The vision system monitors the steering wheel area in order to detect and track drivers' hands and subsequently identify pointing gestures. Intuitive types of gestures other than pointing will also be investigate for use in the system.

The project will also develop algorithms that learn to predict the secondary control that the driver wishes to use next. Such algorithms will be used to adapt the interface to a particular driver, making it quicker and safer to operate.

The project will run from March 1999 until September 2001, and is just one of many on-going projects around the world focussing on the area of intelligent vehicles.

For more details of the ACTIVE project, see the public parts of the project proposal.