Vodafone Procure & Connect, SAP and Accenture are trying out the use of humanoid robots in warehouse environments in order to demonstrate how physical AI could enhance operational efficiency and safety and establish new ways of designing workforces and business models.
Accenture, SAP and Vodafone Procure & Connect will be demonstrating their work at Hannover Messe 2026.
The initiative is an example of the focus of Accenture in terms of applying advanced robotics along with physical AI in real-world industrial settings, assisting organisations transition from experimenting to practical implementation at scale. It also looks at how human-like robots can help evolve future workforce models and at the same time unlock new revenue opportunities throughout various industries.
It is well to be noted that the pilot project of humanoid robots in warehouse environments was implemented at the warehouse of Vodafone Procure & Connect in Duisburg, Germany, where humanoid robots were set up to work with existing warehouse systems.
The robot, apparently, was programmed to perform inspection tasks using the SAP Extended Warehouse Management system and was able to conduct visual inspections autonomously throughout the facility.
During the humanoid robots in warehouse environments, the humanoid robot identified errors, safety risks and optimization opportunities in warehouse processes. It detected misplaced or damaged products, evaluated pallet stacking and weight distribution, identified unused storage space and pinpointed potential hazards like obstacles within the aisles or misaligned pallets. The robot reported its results and suggestions directly into the SAP system, providing real-time visibility and better-informed operational decision-making.
While SAP was responsible for integrating the robots into the warehouse management system, Accenture brought its knowledge in physical AI and advanced robotics as well as digital twin environments in order to design and implement the robot intelligence and operational framework.
Advanced Robotics lead, Accenture, Christian Souche stated that “Trained in digital twins and powered by physical AI, humanoid robots can reduce worker injuries and other warehouse safety incidents and lower overtime costs and the dependency on temporary labour. Equally important, Vodafone Procure & Connect will gather valuable data and insights on robot deployment and performance as a basis for a future humanoid workforce solutions business.”
Head of Embodied AI & Robotics, SAP, Dr Lukasz Ostrowski remarked that “At Vodafone Procure & Connect, we’re leveraging Joule, SAP’s AI execution fabric and interface for embodied AI, connecting robots to end-to-end processes and business logic and enabling them to know why, when and how to act. By grounding actions in trusted SAP data, we can automate health and safety incident reporting and real-time inventory validation to protect workers and strengthen compliance through consistent auditable workflows.”
The global network logistics director at Vodafone Procure & Connect, Reinhard Stefan Plaza Bartsch, said, “Through this pilot, we are exploring how humanoid robotics can improve efficiency, safety and operational visibility in our warehouse operations. It also gives us a clearer view of how these capabilities could scale across our supply chain and support future business models.”
Prasad Satyavolu, Accenture’s global lead for manufacturing, operations and physical AI, opined that “Our work in collaboration with SAP is a great example of how holistic deployment of humanoid robots – from simulation and training to warehouse deployment and integration with SAP data – creates a closed loop with transactional systems.”
The pilot’s humanoid robots are powered by Accenture’s Robot Brain solution that enables them to interact instinctively with human operators via voice and gestures as well as text inputs. They are trained in digital twins of warehouse environments that are built on the Physical AI Orchestrator of Accenture, which makes use of the Mega NVIDIA Omniverse Blueprint and the NVIDIA Metropolis libraries and Blueprint for video search and summarisation for the implementation of visual AI agents so as to go beyond the single repetitive functions and, at the same time, learn new skills by means of imitation and reinforcement learning using the NVIDIA Omniverse libraries.
