Automatic Accident Detection: There’s an App for That
A smartphone app capable of reducing traffic jams by up to 40% has been designed in a joint project by University of Bologna (eng.unibo.it) researchers and Toyota (toyota.com).
“Basically what we’re doing is placing cars in peer-to-peer communication,” says Bologna researcher Marco Roccetti. The app works with an on-board accelerometer, sensor, which detects sudden changes in speed such as that caused by a collision. If an accident occurs, it sends a signal to the smartphone, which in turn instantly sends an accident alarm warning message to cars 300 to 1,000 meters away. These cars will then relay the warning signal to vehicles that are further away from the collision. This allows drivers receiving the warning to avoid the accident area, thereby helping reduce traffic jams.
The Touch-Screen Steering Wheel
While plenty of cars have steering-wheel mounted controls for a variety of functions, researchers at the University of Stuttgart (uni-stuttgart.de), the University of Duisburg-Essen (uni-due.de) and the German Research Center for Artificial Intelligence (dfki.de) are working on a system that is, in effect, like what the iPhone interface is like compared with a BlackBerry.
The steering wheel is made of an 11-mm thick acrylic and ringed with infrared LEDs. An infrared camera attaches to the bottom of the steering wheel to detect reflections when the screen is touched. Trial runs showed that drivers who used the wheel kept their attention on the road up to 77% more than those who use conventional steering wheels.
The researchers are currently working to bring the wheel to market.
Making the Owner’s Manual Obsolete
Researchers at the Technische Universitaet Muenchen Institute of Business Informatics (TUM; tum.de) and Audi AG (audi.com) have developed a virtual dashboard assistant designed to answer a driver’s questions about their vehicle, while providing images and pictures to support relevant details. They call it the “Avatar-based Virtual Co-Driver System”, or AviCoS, and it’s designed to give drivers immediate information about their vehicle, rather than have them page through an owner’s manual.
The virtual avatar is designed to work with the multimedia interfaces that come standard in present Audi vehicles. It’s programmed to understand complete sentences and then answer the driver. There’s even touch-and-tell mode, where drivers can receive more detailed explanations on vehicle functions by touching relevant areas of the screen.
The researchers say their next step is to make AviCoS into a system that recognizes and adapts to a driver’s state of mind. For example, TUM researchers envision the system reducing its multimodal output if it detects the driver is stressed.
The Technische Universitaet Muenchen and Audi have developed a virtual dashboard assistant, AviCoS, designed to interact with drivers to provide information about their vehicle.