South Korea has selected Doosan Robotics to lead two government-backed projects with a combined research budget of about $74 million, spanning AI-enabled collaborative robots and automated welding. The plans put locally processed AI into industrial machines, with commercial robot products targeted for 2031.
Watch Desk analysis
What happened
The AI Insider reports that about $51 million of the combined budget will come from government funding. One project aims to build collaborative robots that use an AI chip mounted on the robot to process camera, touch, force and voice inputs without relying on an external computer or cloud service. Doosan plans three robot models and an industrial humanoid, with 72-hour demonstrations at customer sites.
The second project will use collaborative robots, vision systems and imitation learning to automate difficult welding. Doosan’s targets include welding paths more than 95 per cent similar to those made by skilled workers, with errors of no more than 3 millimetres, and cutting processing time by more than half. Testing is planned at Doosan Enerbility’s nuclear power centre.
Why it matters
This is a substantial attempt to move physical AI from the demonstration floor into work involving manufacturing equipment and safety requirements. The first project’s focus on on-device processing could let robots make decisions locally rather than depend on a cloud connection. The welding project, meanwhile, sets measurable performance goals for work that is difficult to automate.
Those figures are targets, not demonstrated results. Commercialisation of the collaborative robots is planned from 2031; the welding technology would need relevant nuclear-quality certification before Doosan extends it to other sectors.
Our read
The useful detail is the combination of defined industrial tasks, funding and test plans, not the phrase “physical AI”, which is having a busy year. The projects are ambitious, but the real test will be whether the systems meet their stated targets in continuous demonstrations and demanding production environments.
What to watch
- Whether the planned 72-hour robot demonstrations lead to customer deployments.
- Whether Doosan reports progress towards its welding accuracy and processing-time targets.
- Whether the robots reach commercial production on the planned timetable.
- Whether the welding system secures the certification needed for nuclear equipment work.
Discussion spark: Should manufacturers prioritise robots that make decisions locally, even if that means taking on more responsibility for testing and certification?
Sources and evidence
- Doosan Robotics Selected for Two South Korean Physical AI Projects With Nearly $74M R&D Budget (6 October 2026, 12:30 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.