NextFin News - The boundary between enterprise software and physical labor has shifted as the HMND 01 humanoid robot successfully completed a series of autonomous logistics trials at an active automotive production facility. Developed by the UK-based robotics firm Humanoid, the "Alpha Wheeled" variant of the HMND 01 operated within the factory of Martur Fompak, a major automotive supplier, executing complex fulfillment tasks without human intervention. Unlike the choreographed demonstrations often seen at tech trade shows, this pilot took place in a live production environment, integrated directly with SAP’s enterprise resource planning systems.
During the trial, which spanned January and February 2026, the robot functioned as a physical extension of the factory’s digital nervous system. Using Humanoid’s KinetIQ AI platform and the Joule agent layer, the HMND 01 received real-time instructions from Martur Fompak’s operating software. It autonomously navigated the warehouse floor, identified specific pallets, retrieved KLT-format boxes weighing up to 17.6 pounds, and delivered them to designated trolleys. This seamless connection to SAP Business AI allowed the robot to adapt to shifting production demands, effectively treating the humanoid as an "embodied agent" within the broader corporate IT infrastructure.
Artem Sokolov, founder and CEO of Humanoid, characterized the test as a departure from laboratory-bound robotics. Sokolov, who has long advocated for "utilitarian robotics" over aesthetic humanoids, noted that the success of the HMND 01 lies in its ability to be measured against operational standards rather than technical novelty. His stance reflects a growing trend among European robotics startups to prioritize functional integration over the "uncanny valley" realism pursued by some Silicon Valley competitors. However, Sokolov’s bullishness on rapid deployment is viewed with caution by some industry veterans who argue that the wheeled base of the HMND 01 simplifies the mobility challenge, leaving the more difficult problem of bipedal navigation in cluttered spaces largely unaddressed.
The involvement of SAP SE’s embodied AI and robotics group, led by Dr. Lukasz Ostrowski, suggests a strategic pivot for the software giant. By providing the "business context awareness" that robots typically lack, SAP is positioning itself as the essential middle layer between the cloud and the factory floor. Ostrowski’s team is betting that the future of manufacturing lies not in isolated automation, but in a flexible fleet of robots that can be instantly reallocated via software. This approach could significantly lower the barrier to entry for mid-sized manufacturers who cannot afford the rigid, multi-million dollar fixed-automation systems used by the likes of Tesla or Volkswagen.
Despite the technical success, the trial highlights a looming tension in the labor market. While the HMND 01 handled relatively light loads of 8 kilograms, the consistency of its performance under real-world conditions suggests that more strenuous tasks are within reach. Critics of rapid humanoid adoption, including several European labor unions, have expressed skepticism regarding the "collaborative" nature of these machines. They argue that as robots like the HMND 01 become more deeply integrated into SAP-driven workflows, the "human-in-the-loop" becomes a bottleneck that companies will eventually seek to eliminate entirely.
The path to full-scale deployment remains fraught with logistical and financial hurdles. While the proof of concept at Martur Fompak demonstrated that the HMND 01 can follow orders, the cost-per-hour of operating such a sophisticated machine remains high compared to traditional automated guided vehicles (AGVs). Furthermore, the reliance on a constant internet connection for the Joule agent layer introduces cybersecurity risks that many manufacturers are still hesitant to accept. Partners in the project are now moving to assess the pilot results, with plans to explore more complex workflow scenarios that could eventually see the HMND 01 moving beyond simple "pick and place" tasks into more intricate assembly roles.
Explore more exclusive insights at nextfin.ai.
