Bimbo Bakeries USA orders 5 Motiv all-electric EPIC-equipped step vans; delivery in Q3 2019
Air Liquide and Houpu form JV to develop hydrogen station infrastructure in China

TRI-AD, Maxar Technologies, NTT DATA collaborate to build high-definition maps for autonomous vehicles from space

Toyota Research Institute-Advanced Development, Inc. (TRI-AD), Toyota’s automated driving software development company; Maxar Technologies Inc., a global technology innovator powering the new space economy; and NTT DATA Corporation, a leading IT services provider, will collaborate on a proof-of-concept building automated high-definition (HD) maps for autonomous vehicles using high-resolution satellite imagery.


Example of high definition map for automated driving from satellite imagery.

This is an important move toward advancing TRI-AD’s open software platform concept known as Automated Mapping Platform (AMP) and help realize the scalability of autonomous driving.

Autonomous vehicles use several real-time sensors to ensure safe driving and these sensors need to be cross-referenced with an HD map for safe operation. According to TRI-AD analysis, currently HD maps cover less than 1% of the global road network, and there is a need to broaden the coverage of urban areas and local roads before autonomous vehicles can become a mainstream mobility technology.

An HD map created from the accurate satellite imagery allows the driving software to compare multiple data sources and signal the car to take action to stay safe.

In this proof of concept, the three companies will work together to process satellite imagery into vehicle-friendly HD maps. Leveraging Maxar’s cloud-based Geospatial Big Data platform (GBDX), imagery from Maxar’s optical satellite imagery library will be fed into NTT DATA’s specialized algorithms using Artificial Intelligence to extract information that is necessary to generate the detailed road network.

Based on that information, TRI-AD will make HD maps available for delivery from TRI-AD’s cloud into Toyota test vehicles. The group is focusing first on creating an automated HD map for a predefined area of the Tokyo metropolitan region, opening up the future possibility of supporting automated mobility on all roads.

Recent advances in electronics and aerospace engineering are leading to higher resolutions and more frequent updates of global imagery from space-based assets. Additionally, machine learning is helping automate the discovery and integration of semantic relationships between road elements within image data. We’re excited to collaborate with Maxar and NTT DATA to revolutionize automated driving mobility for all.

—Mandali Khalesi, Vice President Automated Driving at TRI-AD



Creating and continuously updating highly accurate road/street maps from space imagery data is or will soon become a reality but it is not an easy task.

To continuously update vehicles precise position using road/street maps data to the accuracy required to keep vehicles at center of driving lane, within a few mms or centimeters, may require further refinement and testing.

Obstructions from tall buildings, clouds, smoke, rain or snow may have to be compensated with added pseudo positioning data. It may be costly but it is possible to do.


Musk is doing it right. His neural network controlled autopilot will make cars smart enough to go anywhere. Things are constantly changing in the real world. Bridges wash out. Landslides along mountain roads. Construction sites appear, accidents happen. Any car that depends on a map from space will run right into all kinds of problems.


Roads/streets proper lane guidance will not be enough for ADVs.

Absolute respect of traffic control lights, city's speed and parking restrictions, other vehicles, pedestrians, animals, bicycles, motorbikes, fallen trees, pot holes and other obstructions will have to be fully considered.

Complementary Lidars, Radars and Camera sensors will also play essential roles.

The comments to this entry are closed.