Radar for end-to-end autonomous driving that outputs raw sensor data and includes a generative simulator for training.
Congruent builds radars for end-to-end autonomous systems that output raw sensor data rather than processed point clouds, enabling neural network training directly from raw radar inputs. The radar connects to vehicles and world models via a digital twin and includes a generative simulator for producing synthetic raw radar data. Teams can record real data, augment scenes in their world model, generate synthetic raw data, test edge cases, and evaluate closed-loop autonomy. Congruent is a product of Congruent.