Automated parking development solution

Kickstart development with pre-made ODD taxonomies, simulation maps, and test suites.

Adapt to the long tail of an ODD with tools to customize taxonomies, maps, 3D content, and test cases using a GUI, parameter-based approaches, and natural language.
‍
Scale simulation across an ODD with cloud orchestration.

Develop safer APS by accurately modeling parking systems with multi-sensor simulation of ultrasonics, fisheye and rectilinear cameras, radar, and lidar. 

Reliably develop, test, and validate sensing, perception, or the end-to-end APS stack by using log re-simulation or sensor SIL or HIL.

Efficiently curate training datasets from drive logs with data mining and auto-labeling tools that leverage generative models to enrich log metadata and simplify search.

Complement real datasets with targeted synthetic datasets that are rendered using both classical and generative sensor simulation and feature annotations and a diversity of parking spaces, signs, road markings, and beyond.

Object-level sensors mock out perception so teams can unit test planning, prediction, and controls.

Realistically test planning, prediction, and controls by utilizing both rule-based and data-driven vulnerable-road-user (VRU) and park-in/park-out behaviors, metrics for parking quality, and physically accurate vehicle dynamics.