While the race for commercially viable humanoid robots continues, some developers are applying physical AI and advanced manipulation to existing platforms. Sanctuary Cognitive Systems Corp. today said it has achieved “world-class performance” in a complex wire-plugging task with a global Tier 1 automotive supplier. The result was a 99.5%+ task success rate at a cycle time of 2.54 seconds, validated against the customer’s live production benchmarks. Sanctuary AI said the milestone reflects its evolved strategy, in which it seeks to deploy its physical AI on existing and next-generation industrial robots.
“Physical AI adoption is gated by AI that meets both performance and cycle-time requirements. That’s what customers are seeking, and that’s what we are delivering,” stated Olivia Norton, co-founder and chief technology officer of Sanctuary AI.
“By focusing on a performance-first approach to physical AI models, we’re delivering value to customers today on an AI platform that will also scale to the next generation of general-purpose systems,” she added.
Proof of concept shows that physical AI is ready for production #
Sanctuary AI’s proof of concept with its automotive customer involved a plug-insertion task requiring manipulation of materials that shifted dynamically while moving on a conveyor. The resulting system matched the throughput speed of the customer’s production line and met the industrial benchmark for performance (see video below).
“Manipulating a flexible wire into a moving target on a live conveyor is exactly the kind of contact-rich dexterity problem that has kept tasks like this out of reach for traditional automation,” said Norton. “Solving it required models built around performance from Day 1 with reliability, cycle time, and safety measured against real production benchmarks. That’s the bar physical AI has to clear to matter in enterprise environments, and it’s the bar we’ve now cleared.”
Sanctuary AI asserted that it has one of the world’s most advanced physical AI teams, including robotics and software engineers who have spent years solving problems the industry said couldn’t be solved. The Vancouver, Canada-based company said it is a hub for machine learning, robotics, and physical AI talent.
In addition, Sanctuary AI noted that its intellectual property portfolio, proprietary hydraulic hands, and advanced AI systems position it as a leader in physical AI. It demonstrated zero-shot learning for dexterous in-hand manipulation in April 2026.
Sanctuary AI makes a strategic shift #
Rather than wait for humanoid robots to reach mass commercialization, Sanctuary AI said it chose to deploy its physical AI on existing platforms. The company claimed that this accelerates its ability to solve the labor challenges facing the manufacturing, logistics, and other industries today.
“This hardware-agnostic approach expedites industrial adoption of physical AI, while building the foundation that will support the next generation of intelligent robotic systems, including industrial humanoids,” said Sanctuary AI.
Whether optimizing an established production line or launching a new one, the company said its customers can gain production-ready performance, faster time to value, and a clear path to the next generation of intelligent industrial systems. “Sanctuary AI gives enterprises a practical, scalable solution that delivers results now and grows with the demands of tomorrow’s factory floor,” it said.
Also in AI-driven grasping this month, Festo tested its GripperAI and launched a lightweight pneumatic gripper, and PSYONIC partnered with ABB Robotics to advance dexterous robotic manipulation.