
As tech gets serious, this year’s CES is less about entertainment advances and more about safety and security measures. Hence Toyota will be showing its high-precision road imaging system that will do away with cars being sent out to scan and map road furniture by sharing what it scans between self-driving cars.
On-board cameras and GPS devices will gather road images and send them to data centres where the imagery is pieced together and a highly detailed, accurate road map produced.
We’re not talking detail about shops, junctions and speed limits, but data on kerb heights, dividing-line widths and traffic-light positions. At the moment, the limited road data available is collected infrequently and is expensive to gather. The high incidence of errors due to cameras over 3D lasers is mitigated by the image-matching available from using so many vehicles to map out any given area.
Toyota claims the new system’s error margin for mapping is restricted to a maximum 5cm on a straight road, and plans to include it as a core element of automated production vehicles on the road by 2020.