General Robotics has announced a major evolution of GRID, its robot intelligence platform, which is designed to automate key stages of robot development, onboarding, skill creation and deployment.
The company says the latest version of GRID can significantly reduce the time and specialist expertise required to move robotic systems from development into production environments.
AI-powered robotics development
Built as an agentic platform for physical AI, GRID combines robotics knowledge, AI models, simulations, software, hardware and operational data into a unified system.
The platform uses knowledge graphs to turn robot deployments, completed tasks, model integrations and system failures into structured intelligence that can be reused across future deployments. General Robotics says this allows the platform’s capabilities to improve continuously as more robots and tasks are connected.
According to the company, GRID can now use AI to engineer processes ranging from onboarding a new robot to creating and deploying new robotic skills.
Faster deployment times
General Robotics says the latest developments have significantly reduced several stages of the robotics deployment process:
- Robot onboarding: From around one month to as little as two hours
- Model ingestion: From three days to as little as 20 minutes
- Skill transfer between robot form factors: From three days to as little as 1.5 hours
- New skill creation and deployment: As little as two days
These improvements are intended to help organisations move beyond individual robotics pilots and deploy intelligent automation across production environments more quickly.
Supporting multiple robot manufacturers
GRID is designed to operate across different robot manufacturers and form factors, giving customers greater flexibility when selecting hardware for specific applications.
General Robotics says its customers include major companies operating in automotive manufacturing, ports, energy generation, food and beverage production, as well as government agencies.
The platform’s ecosystem includes Fanuc and Galaxea, with General Robotics working with the manufacturers to accelerate the deployment of AI-enabled robotic systems.
Building a continuous robotics development loop
The latest GRID architecture is designed to automate the broader robotics engineering lifecycle.
When assigned a task, the system determines the skills required, selects appropriate models and techniques, and can create and run simulation environments for training and evaluation. It can then manage deployment, monitor performance and failures, and identify areas requiring further improvement.
This creates a continuous development cycle spanning robot onboarding, skill development, simulation, deployment, real-world evaluation and optimisation.
General Robotics says the approach is aimed at addressing two major challenges facing the robotics industry: the shortage of specialised robotics talent and the fragmented technology stack required to develop and deploy different robotic systems.
