Why a Humanoid? The Definitive Answer | Michael Perry (Persona AI)
Michael Perry, who led commercial strategy at Persona, argues at the Humanoids Summit that humanoids fit complex heavy industry work better than other automation options.
In this talk
- 3:02
Define what better means
Perry says the right technology depends on the problem and what a site needs to optimize. He points to labor shortages across industries such as chipmaking, steel, construction, and energy, and says every U.S. shift assembling large vessels is behind schedule.
- 5:03
Why humanoids fit heavy industry
Welding involves more than following a seam: workers move and configure equipment, manage cables, fit and grind parts, inspect welds, and work in confined spaces. Perry contrasts humanoids with fixed automation, which can be costly and inflexible, and multiple specialized robots, which can lose utilization and be difficult to coordinate. He says a human sized, dexterous robot could use existing tools and workspaces while taking on varied tasks.
- 18:01
Build task skills from human data
Persona’s pipeline combines motion capture, teleoperation, simulation, and data gathered at customer sites. Perry describes adapting the UMI framework and using cameras and an IMU to collect details such as welding speed and weld puddle movement. He says the company has begun testing autonomous welds at customer sites and feeding results back into the pipeline.
- 24:21
Make the system usable by workers
Perry says workers need to guide robots, resolve problems, and understand what the robots are doing. He contrasts this with a specialized robot operation test that required teleoperators to control a robot while hanging upside down, leaving only two of ten applicants able to do the job. He argues that familiar human movements can help industry workers use the technology without robotics expertise.
Briefs summarise talks and interviews; timestamps open the video at that moment. Claims in a talk are the speaker's own.