LIVE
◆BrainCo Shows Brain-Controlled Robot Platform At WAIC 2026 ◆Hyperscale Data Starts Installing OPR-R2 Robots At Michigan AI Facility ◆PrimeBOT Wins Design Awards For A More Personal Kind Of Robot ◆Walden Robotics Launches With $300 Million To Put General-Purpose Robots To Work Now ◆OlloBot Gives OlloNi SS1 A New Look Before Kickstarter Launch ◆PrimeBOT Brings Personal Robotics To UN's AI for Good Summit ◆Addionics Reveals Battery Architecture For Robots That Never Clock Out ◆Humanoid Robots To Combat Elderly Loneliness In New Pilot From Realbotix and Bloom ◆UMA Shows Off Its First Humanoid Robot And A New Way To Teach It ◆Elon Musk Says AI And Optimus Could Bring Universal Excellent Healthcare ◆Elon Musk Says AI+Robots Will Bring Universal High Income, Michael Burry Claims Revolution First ◆Humanoid Robot Market Could Hit $50 Billion By 2035
Industry 3 min read

YY Group Is Turning Cleaning Work Into Humanoid Robot Training Data

Josh Jones · June 9, 2026 RobotRobotsUnitree Robotics
YY Group Is Turning Cleaning Work Into Humanoid Robot Training Data

YY Group is taking a more direct step into humanoid robotics, and it is starting with the kind of work that keeps buildings running every day.

The Singapore-based company announced a commercial humanoid robotics initiative built around Unitree G1 Edu Ultimate B-U4 humanoid robots. YY Group says the robots will be used as part of a larger push to bring autonomous AI into its integrated facility management business, where labor-heavy jobs like cleaning and maintenance are a constant challenge.

The first target is not a flashy robot demo. It is commercial sanitation and facility maintenance in places like shopping malls, hotels and large commercial real estate properties. That makes sense for YY Group, which already works in workforce management and facility services. The company is not trying to invent a new use case from scratch. It is trying to automate work it already understands.

The most interesting part is how YY Group plans to train the robots.

The company says its cleaning personnel will wear proprietary data-collection gear during normal work shifts. That gear will capture real-world workflow data, including movement, spatial interaction, environmental information and operational decision-making. In plain English, human cleaners will help teach the system how facility work actually gets done.

That data will then run through YY Group’s training lab to support imitation learning and simulation-to-real development. The goal is to convert hands-on human expertise into structured automation data that can eventually support robot deployment, software services and new automation revenue lines.

The Unitree G1 Edu Ultimate B-U4 will serve as the hardware base for the effort. YY Group says the robot includes advanced mobility, 3D touch-sensitive robotic hands and an NVIDIA Jetson Orin onboard AI computing platform.

This is where the business story and robot story overlap. Facility management has always depended on large numbers of people doing repetitive, physically demanding work across big spaces. If humanoid robots are going to break into commercial buildings, they need training data from the actual environments where they will operate. A mall after closing, a hotel corridor during turnover, or a busy mixed-use property is very different from a clean robotics lab.

YY Group says the humanoid initiative will connect with its existing YY Circle workforce optimization platform and its 24IFM software platform. The company describes the model as a hybrid system, combining automated workforce deployment with physical robotic execution.

That is probably the realistic near-term path for humanoids in facility work. Robots will not simply replace every worker overnight. More likely, companies will start by collecting data, automating repeatable tasks, and using robots where they can reduce labor pressure without breaking the operation.

YY Group is making a familiar bet with a very practical twist: before humanoid robots can clean the building, they may need to learn from the people who already do.

Join the Discussion

Stay Ahead of
the Machines

Weekly intelligence on AI and robotics. No noise — just the signal.