17 April 2024  |  Press release

Driving with Language: Introducing Wayve’s Multimodal Driving Model LINGO-2

Wayve unveils LINGO-2, the first vision-language-action model tested on public roads. LINGO-2 generates both driving behavior and textual predictions from the same deep learning model, providing a continuous commentary of its driving decisions.

  • Wayve’s LINGO-2 is a closed-loop driving model and the first vision-language-action model tested on public roads.
  • This is the next step in Wayve’s pioneering work incorporating natural language to enhance the explainability of AI models.
  • LINGO-2 unlocks new innovation for autonomous driving and marks a crucial step toward building confidence in the decision-making process of Wayve’s AI driving models.

LONDON Wayve, the leading developer of Embodied AI for assisted and automated driving, unveils LINGO-2, their closed-loop driving model that links vision, language, and action to help explain and determine driving behavior. LINGO-2 opens up a new dimension of control and customization for an autonomous driving experience. LINGO-2 is the first vision-language-action model (VLAM) to be tested on public roads.

VLAMs hold significant promise for building trustworthy autonomous driving technology. Wayve has been at the forefront of applying large language models to autonomous driving and released its LINGO-1 research model, an open-loop driving commentator, in September 2023. Announced today, LINGO-2 is a closed-loop driving model that deeply links language with driving to provide visibility into the AI model’s understanding of a driving scene.  

Wayve’s AI driving models learn to drive off data and experience without hand-coded rules or HD-Maps. LINGO-2 combines a Wayve vision model with an auto-regressive language model (traditionally used to predict the next words in sentences) to predict a driving path and provide commentary on its driving decisions. Tested in Wayve’s neural simulator, Ghost Gym, and on public roads, LINGO-2 gives a strong initial signal of what can be achieved by aligning linguistic explanations and decision-making. This integration opens up a new level of AI explainability and human-machine interaction that can build confidence and trust in the technology and create a more collaborative driving experience.

Alex Kendall, CEO and Co-founder at Wayve: “LINGO-2 is shaping the future of human-vehicle interaction. The Embodied AI we are building will not only automate driving but also create innovative experiences where the driver can interact with the vehicle to gain more confidence and trust in our assisted and autonomous driving systems.”

Jamie Shotton, Chief Scientist at Wayve: “Language offers a tool for us to interact with robots, understand their decisions, train them faster, and control and customize their outputs. At Wayve, we are driven to keep pushing the boundaries of Science in service of building safe and trustworthy autonomous systems that can meet the needs of everyone, everywhere.”

For more information on LINGO-2, please check out Wayve’s blog.

Back to top