Robotic dogs will be let off the leash in 2026 and take over routine operations – in a recent insight, Siemens describes how the four-legged autonomous mobile robot (AMR) ANYmal from ANYbotics, in combination with COMOS Mobile Worker and the Roboverse Reply platform, will take over inspections in complex industrial plants.
The process chain is crucial: data from ERP, engineering and maintenance systems is consolidated in COMOS Mobile Worker, while Roboverse Reply orchestrates the robotics and filters the feedback so that only relevant information ends up in the interface. This turns robot telemetry into a usable maintenance workflow instead of a separate ‘robotics silo’.
Inspection in multi-storey facilities and under difficult lighting conditions
Technologically, Siemens relies on LiDAR-supported 3D detection and AI-supported mobility, enabling the inspection robots to navigate their way around multi-storey facilities and under difficult lighting conditions. ANYmal autonomously collects data from hundreds of inspection points on demand. The platform can thus visualise anomalies and feed results back into existing systems. This addresses two bottlenecks at once: the shortage of skilled workers in field service and the need to reduce downtime through condition-based measures instead of rigid maintenance intervals.
Explosion-proof quadrupeds and tracked vehicles
At the same time, a market is emerging for robotics operations – software layers that integrate heterogeneous robot fleets, sensors and analytics into the OT/IT stack. A current example is the partnership between Yokogawa and ANYbotics announced on 10 February: Yokogawa’s OpreX Robot Management Core is to be integrated into the ANYbotics software stack to centrally manage the ANYmal fleet. ANYmal X is particularly relevant for process industries. It is positioned as an explosion-proof, ATEX and IECEx Zone 1-certified, walking inspection robot and is scheduled to be launched on the market later this year. The explosion protection trend is not limited to quadrupeds. Mitsubishi Heavy Industries is developing the EX ROVR ‘ASCENT’ explosion-proof plant inspection robot, which runs on tracks and includes a charging/docking station. It is also ATEX and IECEx Zone 1 certified and, with its integrated 6-axis manipulator, is designed to set off on autonomous tours and carry out necessary actions. For operators, this is a sign that a separate ecosystem of certified platforms, sensor payloads and fleet management is establishing itself in hazardous areas.
Decisive and energetic like Lassie
Companies around the globe are driving forward a closed control loop of robotics and AI. This includes Boston Dynamics with its four-legged robot Spot. In the context of a current collaboration with IFS, the leading Swedish provider of industrial AI software, it has been described how its sensor data is to flow directly into an agentic AI that performs analyses and initiates actions – right through to prioritisation and connection to enterprise systems. This shifts the benefit from ‘robots collect data’ to ‘systems decide and initiate actions’ – and makes integration, auditability and governance the real differentiators.
Organisationally, what is needed above all is ‘Robot Ops’
Strategically, this fits into a broader wave of automation: ABB is targeting the growing mid-market with new, more accessible robotics families, while Boston Dynamics is simultaneously preparing humanoid platforms such as Atlas for industrial tasks. For industrial users, this results in a clear prioritisation: value is created when data quality and the semantics of the inspection points are stable and findings automatically flow into work management processes. At the same time, software architecture becomes a key issue because COMOS/OpreX/IFS-like layers determine how quickly use cases scale, how cybersecurity is implemented and how strong the dependence on individual robotics OEMs is. In terms of organisation, what is needed above all is ‘robot ops’ – mission design, data validation and model maintenance are the keywords here – so that autonomous inspection does not fail due to exceptional cases or unclear responsibilities.
https://www.hannovermesse.de/en/news/news-articles/the-humanoid-s-best-friend-??utm_medium=email&utm_source=newsletter&utm_campaign=robo-briefing32-en&utm_term=besucher
