Nomagic's Shoebox Picker handles over 98% of shoeboxes on the market, according to the Polish startup that recently secured funding to deploy physical AI in warehouses. The system represents a shift from single-task automation to adaptable robots powered by foundation models.
"Our vision is to bring physical AI into the heart of warehouse and logistics operations, where intelligent, autonomous systems can finally bridge the gap between digital optimization and real-world execution," said Kacper Nowicki, Nomagic's founder.
Nuro is conducting autonomous on-road testing as part of validation protocols developed through commercial deployments. The company's approach reflects how robotics firms are moving from controlled environments to open-world operation, a transition enabled by AI models that handle edge cases.
The AGV manufacturing market is expanding as foundation models reduce the engineering required for each deployment. Companies no longer need to program robots for every scenario—the models learn from data across installations.
Chinese manufacturers are accelerating adoption by lowering hardware costs. In Saudi Arabia, Chinese robots support logistics, smart manufacturing, healthcare, and smart city services. "They allow local companies and government entities to experiment, pilot, and scale automation solutions in months instead of years, which is exactly what Saudi Vision 2030 requires," said Mohammed Alsolami, a regional observer.
Alsolami added: "I believe Chinese robotics is playing a clear role in narrowing the technology gap globally."
Talent movement signals industry maturation. Engineers from OpenAI and Apple are joining robotics startups, bringing expertise in training large models and deploying AI products at scale. This migration follows investment patterns that favor physical AI over pure software plays.
The convergence of foundation models, falling hardware costs, and commercial validation is turning robotics into a platform technology. Systems that once required months of custom development now adapt within weeks. Covariant and other logistics-focused companies are deploying similar capabilities across warehouses, while autonomous vehicle operators test in public environments.
The transformation spans industrial and consumer segments, with applications ranging from fulfillment centers to last-mile delivery. As models improve and hardware costs drop, robotics companies are targeting general-purpose systems rather than specialized machines.