Inbound Logistics | January 2026

AI Powers Next-Gen Robotics As robots shift from rigid, rule-based systems toward more flexible and intelligent operations, AI is unlocking new levels of autonomy, perception, and adaptability across robotic form factors—becoming the central nervous system for next-generation robotics. CLOSEDLOOP CONTROL AND PERCEPTION Physical AI is driven by real-time sensor fusion, machine vision, and embedded inference. Physical AI enables robots to react dynamically to their environment. Cobots, for instance, now use closed-loop systems that continuously ingest and process sensor data to adapt movements on the fly. This enables them to function safely alongside humans in unpredictable settings. AI AGENTS FOR SPECIALIZED TASKS Lightweight, domain-specific agents are emerging to give robots skills such as palletizing, pick-and-place, or machine tending. These agents often live at the edge or within robot controllers and support integration into workflows with minimal training data. ROBOTICS FOUNDATION MODELS AND WORLD MODELS The most transformative trend is the rise of robotics foundation models, which allow robots to generalize

AutoStore deployments

across unseen tasks, objects, and environments. Vendors like Dexterity, Skild AI, and NVIDIA have demonstrated early commercial traction, while players like Google DeepMind and Meta continue development in research labs. These models combine reinforcement learning with diverse datasets to reduce time- to-deployment and scale robot capabilities across use cases. Closely linked to foundation models are world models, a more advanced AI architecture that allows robots to simulate their environment, predict object behavior, and “think ahead.” World models enable model-based autonomy for complex robots like humanoids.

The Balluff project consisted of an AutoStore grid with 20,110 bins, 7 robots, 1 ConveyorPort, and 3 CarouselPorts. The installation was able to maximize space utilization within the existing infrastructure. FASHIONING A COMPACT SOLUTION Bleckmann, a supply chain management provider for fashion and lifestyle brands, went live with its AutoStore implementation at its distribution center in Bergen op Zoom (Netherlands) in December 2025. The AutoStore system uses up to 7 times less space while dramatically increasing storage capacity. The system minimizes picking errors, improving quality and efciency. This enables multiple customers’ SKUs to be stored simultaneously, reducing stock shortages and optimizing fulllment. Radio-controlled robots retrieve items, eliminating manual picking and accelerating processing. This solution supports later cut-off times for next-day delivery, demand forecasting, and rapid dispatch during peak periods. The AutoStore system also delivers sustainability benets. Ten robots consume the same amount of energy as a household vacuum cleaner and can operate in the dark.

Boston Dynamics’ Atlas humanoids may soon integrate Gemini Robotics AI foundation models, thanks to a new partnership between the robot maker and Google DeepMind. (Photo: Boston Dynamics)

Source: ABI Research

164 Inbound Logistics • January 2026