We are advancing our development guided by three core technical beliefs.
-
POLICY
AI will revolutionize many fields and will ultimately achieve autonomous driving.
-
POLICY
Simplicity triumphs over complexity, and End-to-End (E2E) is the key.
-
POLICY
A Generative AI that understands the world is the only way to achieve fully autonomous driving.
AI will revolutionize many fields and will ultimately achieve autonomous driving.
AI technology has rapidly evolved over the past few years, and autonomous driving innovations brought about by AI are progressing at an unprecedented pace. Our first technical belief is that AI will understand traffic systems and outperform human drivers due to learning capabilities that allow it to adapt to complex scenarios and push beyond existing limitations. This transformation goes beyond technological advancement to make mobility safer and more efficient for society.
Simplicity triumphs over complexity, and End-to-End (E2E) is the key.
A simple yet powerful End-to-End (E2E) AI developmental approach will be the most efficient and effective solution. Traditional autonomous driving systems are often entangled in numerous complex modules, making overall optimization a challenge. The E2E approach maintains simplicity across the entire system for outstanding performance. This is crucial in evolving technology, as it facilitates system maintenance and enhances scalability.
A Generative AI that understands the world is the only way to achieve fully autonomous driving.
The key to fully autonomous driving lies in a generative AI that can understand the world and make predictions. Just as human drivers comprehend the world around them and make decisions based on that understanding, our generative AI learns from vast amounts of data to make optimal real-time decisions. Our generative AI is designed to respond to situations on the road and to unpredictable events. This allows us to address long-tail scenarios and take solid steps toward fully autonomous driving.