Robot dogs are now doing the night shift at factories, and software just started supervising them

ANYbotics renamed its Data Navigator software to Shift, an operations platform built to manage fleets of its ANYmal four-legged inspection robots. Shift schedules inspection missions, tracks robot status, and supervises the work as robots move through industrial facilities around the clock.

Why it matters: this isn't a cosmetic rebrand. ANYbotics says the name reflects a real change in what the software does, moving from giving customers access to one robot's data to coordinating entire fleets running continuous plant operations.

Background: Data Navigator originally helped customers access and navigate inspection data collected by individual robots. As ANYbotics customers scaled from single robots to full fleets handling daily plant operations, the software had to grow with them.

Jacob Euler-Rolle, ANYbotics' Head of Product Management for Workflow Integration, put it plainly: "Robots now run their own shift around the clock. Shift schedules and supervises that work, then hands findings and evidence to the team" each morning.

In real life it's like a night-shift supervisor who never needs a coffee break, except the workers it's supervising are robot dogs walking past pressure gauges instead of people on a factory floor.

By the numbers: ANYbotics cites estimates that undetected equipment failures cost major industrial companies roughly $1.4 trillion annually in downtime, with each hour of an outage averaging about $250,000.

Why it matters: a robot spotting a problem is only useful if someone acts on it quickly. Shift is built specifically to close that gap, turning a robot's sensor reading into a work order inside a plant's existing maintenance software, rather than leaving it sitting in a report nobody reads until it's too late.

Shift covers four areas. Fleet management handles scheduling and monitoring, including battery levels, positioning, and live video from each robot. Fast deployment tools help teams extend inspection coverage to new sites quickly. Asset intelligence analyzes thermal, visual, and acoustic sensor data over time to catch meaningful changes early. Workflow integration sends confirmed findings into systems like SAP, GE APM, and Microsoft Teams, so the data lands where maintenance teams already work.

Why it matters: each piece addresses a different failure point in getting from "robot sees something" to "human fixes something." Splitting it into four integrated modules suggests ANYbotics learned which specific steps customers were getting stuck on.

A US aluminum manufacturer running three ANYmal robots in parallel, coordinated through Shift and feeding results into its SAP workflows, estimates it has avoided approximately $3 million in downtime. A metals operation under the Viohalco Group has also used the platform to get a single view of asset condition across steel, aluminum, and copper plants.

Why it matters: industrial robotics pitches are full of theoretical savings. A customer citing a specific dollar figure tied to a real deployment is a more concrete signal than a vendor's own projections.

Not currently. The platform is built around ANYbotics' own ANYmal and ANYmal X robots. But the company has described a longer-term goal of extending Shift into a hardware-agnostic command center that could eventually manage third-party robots too.

Why it matters: if that vision plays out, Shift would stop being just a tool for ANYbotics' own hardware and start competing as management software for industrial robot fleets generally, a much bigger market than one company's own robot sales.

Background: ANYbotics was founded in 2016 as a spin-off from the Robotic Systems Lab at ETH Zurich, one of the world's leading robotics research institutions. The lab had been building a series of quadruped research robots since around 2009, and ANYmal grew out of that decade of research before becoming a commercial product.

Why it matters: ANYmal wasn't built from scratch as a product pitch. It emerged from years of academic research into legged locomotion, the kind of work that later became influential across the broader field of learning-based robot movement. That research pedigree is part of why ANYmal can reliably climb stairs and navigate rough industrial terrain that would stop a wheeled robot.

Background: ANYbotics isn't alone in selling legged robots for industrial inspection. Boston Dynamics' Spot robot competes in a similar space, and both companies have pushed into oil and gas, chemical, and power generation facilities where hazardous or hard-to-reach equipment makes human inspection slow or risky.

Why it matters: as more companies field competing robot fleets, the software managing those fleets becomes just as important competitively as the robots themselves. ANYbotics positioning Shift as eventually hardware-agnostic suggests it sees fleet management, not just robot sales, as where the next phase of competition will play out.

Background: ANYbotics has also been expanding what ANYmal can physically access. The company's ANYmal X variant is designed to be Ex-certified, meaning it's rated safe to operate in hazardous, potentially explosive environments like oil and gas facilities, a certification standard that most robots don't meet.

Separate recent updates have added automated door access, letting robots pass through both unlocked and locked doors without a person present, and gas-leak detection payloads that can locate leaks and measure concentrations.

Why it matters: each of those capabilities removes a specific reason a human still needed to be involved. A robot that can't open a secure door, or can't legally enter a hazardous zone, still needs a person to bridge that gap. Shift's fleet-wide scheduling only becomes fully "around the clock" once the robots themselves can handle those physical obstacles on their own.

ANYbotics says current deployments won't be disrupted by the change. Customer data, configurations, logins, and API endpoints all stay the same, and the Shift 1.0 interface updates automatically with no changes to pricing, packaging, or contracts.

• If your industry relies on physical inspections (energy, mining, manufacturing), robot fleets managed by software like this are moving from pilot projects to standard operations.
• The real value isn't the robot walking around, it's what happens after it spots something. Shift's bet is that routing findings directly into existing maintenance software is what actually prevents downtime.
• Reported dollar savings, like the $3 million downtime figure here, are becoming the real benchmark for industrial robotics, not just how advanced the robot itself looks.
• Watch for "hardware-agnostic" fleet management platforms becoming their own competitive category, separate from who makes the robots themselves.