Story perspectives

NYU Develops BrainBody-LLM: Boosts Robot Task Success by 17%

12/1/2025

Summary

Story summary

Researchers at NYU Tandon School of Engineering developed BrainBody-LLM, enabling robots to plan actions and learn from feedback.
The system uses large language models to translate tasks into robotic commands.
Testing in simulations and on a robot yielded a 17% increase in task completion rates.
The approach is not yet ready for widespread use, tested only in controlled environments.
Future work will add diverse sensor modalities to improve real-world applicability.

1 of 1

Story summary

Researchers at NYU Tandon School of Engineering developed BrainBody-LLM, enabling robots to plan actions and learn from feedback.
The system uses large language models to translate tasks into robotic commands.
Testing in simulations and on a robot yielded a 17% increase in task completion rates.
The approach is not yet ready for widespread use, tested only in controlled environments.
Future work will add diverse sensor modalities to improve real-world applicability.