Case Study: Toyota Research Institute achieves scalable, high-quality sensor-fusion training data with Scale AI

Toyota Research Institute - Customer Case Study

Toyota Research Institute (TRI), which develops automated driving modes Guardian and Chauffeur, faced large volumes of sensor data but limited ability to label it at the scale and quality their safety‑critical research required. To solve this, TRI engaged Scale AI, using services such as Sensor Fusion Cuboids, Sensor Fusion Segmentation, and Semantic Segmentation to meet strict annotation and quality needs.

Scale AI implemented multi‑modal annotation pipelines (2D and 3D bounding boxes, semantic and sensor‑fusion segmentation), ran TRI in the private beta for Sensor Fusion Segmentation, and provided flexible workflows and rapid feature delivery. As a result, Scale AI ramped throughput 10× in weeks, supported four large annotation pipelines without significantly increasing TRI’s team size, and turned around features in as little as 24 hours—enabling faster research and continued use of Scale AI’s latest annotation capabilities.

Toyota Research Institute

David Garber

Product Manager