ANYbotics Shift Turns a Robot's 2 a.m. Gas Alert Into a Work Order Before Dawn
A new platform connects walking industrial robots to SAP, GE Vernova, and dozens of other enterprise tools, so a fault spotted overnight becomes a repair ticket before the morning shift arrives.
Key points
- ANYbotics launched Shift, a cloud-based industrial inspection platform, connecting its ANYmal four-legged robots to enterprise software including SAP and IBM Maximo.
- Undetected equipment failures cost the world's 500 largest companies an estimated $1.4 trillion a year, with one hour of downtime averaging $250,000, according to ANYbotics.
- Shift's four modules cover fleet scheduling, sensor-based anomaly detection, automatic work-order creation, and live plant mapping.
- Training takes days rather than weeks; ANYbotics Academy had already trained more than 1,000 users at launch.
- Full fleet management for robots from other manufacturers is planned for 2027.
Picture a four-legged robot the size of a large dog walking a chemical plant at midnight, scanning motors with a thermal camera, listening for a gas leak. That part exists already. What's new is what happens next: the robot's findings skip every email chain, land directly inside the plant's maintenance software, and become a repair ticket before the morning shift starts.
That's the pitch behind Shift, announced this week by ANYbotics, the Zurich-based company founded in 2016 that makes the ANYmal quadruped robot. As first reported by The Robot Report, Shift ties the robot's sensory readings to the software stacks that industrial plants already run.
What does Shift actually do?
Four separate modules, each handling a distinct job. Shift Fleet covers scheduling and monitoring across an entire robot group, letting a control system trigger an unplanned inspection automatically when a process reading looks odd. Shift Insight layers operating context on top of raw sensor data: a thermal reading only counts as an overheating alert when motor load and ambient temperature also support that conclusion, cutting false alarms. Then Shift Connect turns a confirmed anomaly into a work order inside tools like SAP or GE Vernova's asset performance management software, with alerts delivered through email or Microsoft Teams; SMS and WhatsApp support is planned. Shift Maps keeps a live digital model of the plant, anchoring every reading to a specific piece of equipment by its asset ID.
The platform is certified to ISO 27001, a widely recognised information-security standard, and offers end-to-end encryption plus an on-premises option for plants with strict data-residency rules.
How hard is it to set up?
Not very. Jacob Euler-Rolle, ANYbotics' head of product management, told The Robot Report that a first deployment needs no enterprise software integration at all. The fleet runs independently straight away. Connecting to existing systems means granting network access and mapping asset IDs; pre-built connectors already cover SAP, GE Vernova, Siemens Energy, Cognite, SLB, Yokogawa, and others through a REST API.
When InOrbit opened its robot fleet software in September, the same challenge was front and centre: the industry doesn't struggle to make robots that move or sense, it struggles to get their data flowing cleanly into the systems that decide what humans do next. ANYbotics is making the same bet.
Should you watch this closely?
Yes, but with one eye on the trust problem. The anomaly-to-work-order pipeline is only useful if engineers believe it enough to act without double-checking every alert. False positives will erode that confidence fast, and no amount of pre-built connectors fixes a reputation for crying wolf. The hardware is fine. The calibration of Shift Insight's thresholds is the thing worth watching.
