USE Network launch I UAV Works VALAQ l Cable harnesses l USVs insight l Xponential 2020 update l MARIN AUV l Suter Industries TOA 288 l Vitirover l AI systems l Vtrus ABI

Video feeds combine to produce HD maps Driverless cars Toyota’s automated driving lab has demonstrated the ability to combine video feeds from cameras in cars with satellite imagery to create high definition maps with a resolution of under 50 cm for automated driving systems (writes Nick Flaherty). The proof of concept developed by Toyota Research Institute-Advanced Development (TRI-AD) combined several methods for HD map building, which is an essential part of any driverless car’s control system. The first step is to build the maps from data derived from the cameras of ordinary vehicles as well as satellite imagery, without the use of conventional means of collecting data from sources such as survey vehicles. Data from consumer dashcam cameras was used to detect and place key road features such as lane markings, traffic signals and signs in Tokyo and two cities in the US, which gave a relative accuracy of 40 cm. More map information was automatically extracted from high-resolution satellite data provided by Maxar Technologies, removing and correcting non-map image pixels such as automobiles, shadows and occlusions caused by the intrusion of buildings in satellite images. This data was used to improve the accuracy to 25 cm. TRI-AD then used its Automated Mapping Platform to get data from a mapping system run by TomTom by converting data formats and applying correcting algorithms. This showed that maps for urban roads, including lane markings necessary for automated driving, could be successfully created or updated in near real-time. The final stage of the project was with map company HERE Technologies, collecting and correcting the positional errors in the vehicle sensor data. This automatically generated HD maps including lane-level information. Combining satellite images and video feeds boosts the accuracy of maps for automated systems