Better use of existing intelligence drives automotive research
‘Cognitive cars’ that use multicore microprocessors designed for video consoles to connect together a vehicle’s standard task-specific processors could herald a new layer of automotive intelligence.
Ontario’s McMaster University Faculty of Engineering and IBM will collaborate on a project that investigates ways in which disparate automotive processors can be integrated to support new applications, such improved vehicle failure detection, rerouting tools, and smart accident avoidance. McMaster University’s researchers’ work will later look at how such systems can be developed to help reduce driver tension, and emissions-related to ‘stop-and-go’ traffic.
“Research has focused on safety-critical software in the nuclear energy industries and medical devices, but the automotive industry is increasingly adding ‘active safety’ functionality to road vehicles,” said team leader Dr Alan Wassyng, associate professor, department of computing and software, at McMaster University. “Use of powerful multi-core processors allows us to use redundancy techniques more successfully, and provide global instead of local solutions, because a central processor can collect results from other onboard processors/sensors."
It is also powerful enough to handle sophisticated image-processing algorithms for sensing its environment, or the driver’s status, and so on, Dr Wassyng adds.
The McMaster/IBM initiative will also study how increased processing power might enable vehicles to integrate better into regional and global transportation systems, including traffic management, roadside service, and emergency services. The nine-member research team will base their work around IBM’s Rational integrated software development environment.