Understand.ai tools up for autonomous driving

First published at ITS World Congress - October 23, 2019

With the need for high-quality training and validation data for autonomous driving, Understand.ai offers what it says is a world-class toolchain, including 2D and 3D annotation tooling.

This allows annotations for every common annotation type, including 2D bounding boxes, pixel-wise semantic segmentation, tagging and polyline annotation as well as 3D bounding boxes for lidar and radar data.

In addition, Understand.ai’s tool suite has the potential to be licensed as software-as-a-service, plus it offers a one-stop annotation service solution together with trusted partners worldwide.

For validation purposes, Understand.ai provides a scenario reconstruction service from real-world data recordings such as cameras and lida, allowing customers to re-use recorded data directly in their simulation environments to debug their AD-algorithms based on real-world data.

Understand.ai also collects its own data together with fleet partners, using it to extract highly diverse scenarios.

These are stored in a scenario database which customers can license to complement their own data recordings, further helping enlarge the coverage of the enormous test-space applying with automated driving.

Booth 540