Waymo's new model, EMMA, processes sensor data to predict future trajectories for its robotaxis, enhancing decision-making and obstacle avoidance.
This development signals a paradigm shift in autonomous driving as MLLMs may transition from traditional language tasks to complex real-world driving challenges.
Historically, autonomous systems relied on distinct modules for different tasks, leading to inefficiencies and challenges in adapting to new environments.
Waymo's research suggests that MLLMs like Gemini could tackle inter-module communication issues, improving the adaptability and scalability of autonomous driving systems.
Collection
[
|
...
]