-
Brain Corp and Tennant Company are rolling out an AI-powered navigation upgrade to thousands of robotic floor scrubbers already operating in hospitals, airports, and warehouses. The software, called BrainOS Clean 2.0 with SelfPath AI, lets the machines redraw their own cleaning routes on the fly when they encounter a fallen barrier or an unexpected crowd of people. No technician with a tablet. No fresh map to upload. The robot just adapts.
The companies said the upgrade cuts deployment time for a new facility by more than three times, boosts the robots' ability to operate without human intervention by 55 percent, and expands the total floor area each machine can cover by 22 percent, according to RoboticsTomorrow's coverage of the announcement. BrainOS Clean 2.0 begins rolling out the week of April 6 to select Tennant X-series robotic floor cleaners, with full fleetwide availability in May.
The numbers are from Brain Corp and Tennant's own marketing materials and have not been independently verified. But the underlying technology is grounded in real operational history. Brain Corp's fleet of more than 37,000 autonomous mobile robots has logged more than 19 million hours of autonomous operation and covered more than 250 billion square feet of floor space worldwide, the company said in a March milestone post on its website. That's the largest fleet of autonomous mobile robots in the world by unit count, according to Brain Corp.
In a statement, Pinn said SelfPath AI means deployments happen faster, route retraining is eliminated, and cleaning performance improves in dynamic environments. Schottler said customers operate in constantly changing environments, and BrainOS Clean 2.0 lets them deliver a more responsive autonomous solution without replacing a single machine.
The relationship between the two companies runs deeper than a typical vendor agreement. Tennant invested $32 million in Brain Corp as part of a February 2024 exclusive technology agreement, according to Tennant's investor announcement, formalizing a partnership that had been building for years. Brain Corp has raised $193 million over five funding rounds, according to funding tracker Tracxn. In June 2025, Tennant sold its 10,000th autonomous mobile robot, a milestone that underscores how aggressively the 118-year-old cleaning equipment manufacturer has pushed into robotics. The company had sales of $1.20 billion in 2025 and employs roughly 4,500 people.
Tennant's X6 ROVR, a mid-sized robotic scrubber capable of cleaning up to 75,000 square feet per cycle, is among the machines that will receive the SelfPath upgrade. That's roughly the floor area of two and a half Costco warehouses.
The core capability SelfPath claims to deliver is not new in principle. Robotic floor cleaners have had some form of autonomous navigation for years, and the ability to dynamically reroute around obstacles is standard on most modern units. But the practical difference Brain Corp is describing is in scope: rather than a robot following a discrete set of manually taught routes and pausing when something gets in the way, SelfPath is designed to assess the environment continuously and recompute a coverage path across the entire floor, not just reroute around the immediate obstacle.
In a real facility, that distinction matters. A hospital corridor blocked by a gurnee at 9 a.m. looks different from one at 3 a.m. when the gurnee is gone and the supply cart has moved. A warehouse that rearranges its aisles for peak season breaks most robots configured for off-season layouts. The question is whether SelfPath's adaptation is fast enough and reliable enough to handle those transitions without requiring a human to step in and babysit the reroute.
The 55 percent autonomy improvement claim is the one worth watching most closely. "Autonomy" in the context of robotic cleaning is a spectrum. A machine that cleans 90 percent of a floor without intervention is useful. A machine that cleans 99 percent is transformative. The difference between those two numbers is where the human janitorial supervisor either stays on the floor or moves to a monitoring screen. That is a real labor economics question, and it is the one the companies are not yet answering with published data from live deployments.
For now, the story is a software upgrade on proven hardware. Brain Corp and Tennant are betting that their installed base of thousands of machines is a moat: instead of selling new robots, they sell an upgrade to machines already running. If the claimed improvements hold up in real facilities, it is a clean business model. If the autonomy numbers come mostly from controlled demos, customers will notice quickly.
What to watch next: independent customer data, specifically the manual intervention rate before and after the upgrade. That single number will tell you more than any press release.
