About

Robots have always been specialists in fixed work. For decades, industrial robots have welded and painted faster and more precisely than people — but only for jobs where the same items arrive at the same place and are handled with the same motion. Work where every day looks different — different items, different placement, space shared with people — makes up most of Japan's workplaces, and robots could never touch it.

That assumption is starting to change, because the intelligence of generative AI has met the body of the robot. Robot foundation models (VLA) see through cameras, understand language, and move an arm — all with a single AI. Instead of writing programs, robots learn from human motion. Robotics around the world is racing toward its “ChatGPT moment.”

But a clever brain alone can't hold down a job on site. The delicate force it takes to pick up something fragile. The know-how that differs by industry. And deployment fast enough to start working without teaching from scratch. With Muso-roid — industry-specific robot foundation models combined with smart force control — we build physical AI made to work on real sites, not just in demos.

Japan, where the workforce keeps shrinking, can be the first country where physical AI becomes the norm. That future takes shape only when the voices of the field, the passion of engineers, and the wisdom of partners come together. Let's build the “new normal” of the future, together.