Boston Dynamics and Toyota Research Institute (TRI) inked a research pact designed to accelerate developments in humanoid robot technology, using a combination of AI models and the robotics company’s Atlas hardware.

Unveiling the research collaboration, the pair claimed it would combine world-leading teams developing AI systems and robotics, with activities set to include recording data from having Atlas perform various tasks and using it to support the training of TRI’s Large Behaviour Models (LBMs).

The two aim to “demonstrate that large, pre-trained models can enable the rapid acquisition of new robust, dexterous, whole-body skills” for robots and accelerate further development of “general purpose humanoids”.

Boston Dynamics and TRI will also seek to answer “fundamental training questions” around humanoid robots, assess the ability of research models for whole-body sensing, and understand “human-robot interaction and safety/assurance cases to support these new capabilities”.

TRI has been working on various AI developments including LBMs, which are AI models designed to aid robot training. Boston Dynamics is best known for creating its humanoid Atlas robots and dog-form creation Spot.

Boston Dynamics CEO Robert Playter hailed the partnership as “an example of two companies with a strong research and development foundation coming together to work on many complex challenges and build useful robots that solve real-world problems.”

TRI CEO and Toyota chief scientist Gill Pratt added “recent advances in AI and machine learning hold tremendous potential for advancing physical intelligence,” claiming the implementation of its AI with Boston Dynamics’ hardware is “game-changing for each of our organisations as we work to amplify people and improve quality of life”.