Home » Helm.ai unveils new AI simulation tool for Autonomous Driving

Helm.ai unveils new AI simulation tool for Autonomous Driving

Helm.ai, a leading provider of AI software for autonomous driving and robotics, has announced the launch of WorldGen-1, a groundbreaking multi-sensor generative AI foundation model. This innovative tool simulates the entire autonomous vehicle stack, streamlining the development and validation of autonomous driving systems.

WorldGen-1 generates highly realistic sensor and perception data across multiple modalities and perspectives, predicts the behavior of vehicles and agents, and extrapolates sensor data from one modality to another. Trained on thousands of hours of diverse driving data, WorldGen-1 can simulate sensor data for cameras, lidar, and odometry, accurately replicating real-world situations.

This comprehensive sensor simulation capability enables the generation of high-fidelity multi-sensor labeled data, resolving and validating challenging corner cases. WorldGen-1 can also augment existing camera-only datasets into synthetic multi-sensor datasets, reducing data collection costs.

Beyond simulation, WorldGen-1 predicts the behaviors of pedestrians, vehicles, and the ego-vehicle, generating realistic temporal sequences. This enables AI-generation of potential scenarios, including rare corner cases, and demonstrates advanced multi-agent planning and prediction capabilities.

According to Helm.ai’s CEO and Co-Founder, Vladislav Voroninski, WorldGen-1 takes a significant step towards closing the sim-to-real gap for autonomous driving, accelerating development, improving safety, and reducing the gap between simulation and real-world testing.

“Generating data from WorldGen-1 is like creating a vast collection of diverse digital siblings of real-world driving environments,” added Voroninski. “This enables us to tackle the most complex challenges in autonomous driving.”

Source