YYForce robots projected to cut selected costs 10% at 772-room Singapore hotels

YYForce (NASDAQ: YFOR), Novotel Singapore on Stevens and Mercure Singapore on Stevens announced on October 13 that service robots will roll out across the hotels' shared precinct, under a partnership marked with a memorandum of understanding signing. The two hotels have 772 guestrooms between them, plus restaurants and event spaces.

A comparison prepared before deployment estimates that monthly operating costs in selected housekeeping, concierge and food and beverage operations will fall by roughly 10%, and the model already counts what the robots themselves cost. For the hotels, the open question is how much of this is a purchase they can judge today and how much is research they are hosting.

A confirmed rollout and a separate research track

The split of roles is clear. The hotels bring hospitality operating expertise and a working environment where the technology can be evaluated. YYForce supplies the robotics and technical support while it builds what it calls an integrated human, AI and robotics workforce model.

The first phase is confirmed. Specialized service robots will handle guest deliveries, food and beverage logistics, scheduled cleaning and concierge support. The goal is to take repetitive delivery, transport and cleaning work off staff, whom the hotels call Heartists, so they can spend more time with guests.

Even this phase serves two purposes. According to the partners, the rollout gives YYForce a live operating environment for testing robots in hotel logistics, facilities work and guest-facing services. The hotels get help with the trips and the cleaning, and the vendor gets a busy, working precinct to learn from.

The second phase works very differently. It centers on YYForce's humanoid system and later versions of it. Selected housekeeping and banquet workflows will feed into development of the robot's hardware, software and AI, through permitted observations and agreed arrangements for using operational data. In phase one the hotels are customers. In phase two they work more like a live test site.

Piotr Kupiec, general manager of both properties, set the test in operating terms. Success means more consistent service and everyday work that's more manageable for his teams. Running two hotels, he said, gives the partners a chance to rethink how work moves through the whole operation. That measure is wider than the cost line, and it's the one his staff will notice first.

His case starts with the trips themselves. Kupiec said that inside a hotel, each run to drop off an item and each repetitive chore uses up time. He said the goal is to apply robotics with care, freeing staff to put their attention on what guests value most. Set beside the cost model, that makes the delivery trip the core issue. It is the unit of work the robots are supposed to take over, and the savings depend on it.

What sits inside the 10%

Two things about the projection matter. It covers selected housekeeping, concierge and F&B operations, not everything either hotel spends. And it already counts what the robots cost, which makes it a net figure a hotel finance team can check against real invoices once the machines are running.

The three areas in the model line up with the phase-one duties: concierge support, scheduled cleaning for housekeeping, and guest deliveries plus F&B logistics. That match suggests the projection prices the specific jobs the specialized robots will take on, rather than anything the humanoid research might deliver later.

The partners say plainly that actual costs, productivity and guest-service outcomes will be measured after rollout, so the 10% is a hypothesis with a test already attached.

For hotels where a lot of staff time goes into carrying items between kitchens, lifts and guest floors, the Stevens results would be a closer comparison than numbers from a demo. For properties where that kind of running around is a small part of the day, the figure may not carry over.

Lifts, maps and networks come first

Before any robot carries a towel, there's a checklist: lift integration, route mapping, network requirements and guest arrival notifications. Three of those four belong to engineering and IT, not housekeeping.

Kupiec's line about rethinking how work moves through the operation reads almost literally here. Before delivery work can move from people to machines, the paths it travels along, from lift lobby to corridor to guest door, have to be mapped, connected and announced.

Humanoids rehearse in mock rooms first

On October 6, a week before the hotel announcement, YYForce opened a Robotics Training, Data and Experience Center in Singapore, according to its PR Newswire release. The center has mock service spaces for training humanoid robots in selected hotel housekeeping, restroom cleaning, banqueting and hospitality support tasks. The company is collecting data there on navigation, movement, object handling, task execution and environmental conditions.

The mock spaces also test how human workers, humanoid robots and specialized service robots work together. That three-way mix is the same arrangement the Stevens deal is sequencing: people and specialized robots first in the live hotels, humanoids later.

YYForce CEO Mike Fu said the hotel partnership moves the company past controlled demonstrations and toward real-world work. In his view, the collaboration pairs technology that hotels can put in place now with the operating experience needed to build the next generation of workforce solutions. He said the objective is to build "the technology, operational knowledge and permitted training data" for a future in which people, AI, specialized service robots and humanoids work side by side in hospitality, facility management and other service-intensive industries.

The strongest reason not to read this as humanoids arriving in hotel corridors comes from YYForce itself. The company says its training and demonstrations don't show that humanoid robots can yet do complete housekeeping or restroom-cleaning jobs on their own.

What the training is meant to do is help evaluate performance and adapt workflows to specific service environments. A mock room can only stand in for a real one up to a point, which is why observed banquet and housekeeping workflows at two operating hotels could matter more to the humanoid program than anything the robots contribute to the hotels' cost line in the near term.

A staffing company selling robot service

The vendor's background shapes how this could be bought. YYForce describes itself as a technology-enabled provider of workforce solutions and integrated facility management. Staffing Industry Analysts notes that it was formerly YY Group, that its 2025 revenue rose 39.3% to US$57.2 million, and that in 2025 it signed a partnership with service-robot maker Keenon to explore deployments in Southeast Asian hospitality.

Stevens isn't YYForce's first hotel robot, either. On October 5, the same outlet reported that the company, through its YY Circle Malaysia brand, had put a hospitality service robot to work at Aloft by Marriott Langkawi Pantai Tengah, a Malaysian property, to help staff with routine tasks. YYForce said it would monitor task completion, reliability, staff adoption and guest feedback there to shape future deployments.

Hotels elsewhere have gone further: robots are already used across hotels in Japan, and China plans a fully robot-operated hotel. YYForce is pitching the opposite end of that range, robots inside staffed hotels, which fits a company whose core business is supplying people.

At the Singapore center, YYForce says it is evaluating robotics-as-a-service, leasing and managed robotic services, which would let customers use robots without buying the hardware outright. It's also looking at integrated contracts that combine people and robots in one service. Deployments will be judged on task performance, safety, reliability, how much human supervision is needed and customer economics. The Stevens release doesn't describe the commercial terms for these two hotels.

The arrangements under study could bundle robot access with deployment support, maintenance, monitoring and human service capabilities. For a hotel, that structure would move upkeep and oversight of the machines into a service contract, which is familiar territory for anyone who already buys outsourced cleaning or staffing.

The center also lays out the route YYForce expects deals to take: training and data collection, then workflow testing, customer demonstrations, proofs of concept and, where viable, commercial deployment. Hotels sit on its list of target settings alongside commercial buildings, residential developments, retail premises and warehouses. On that map, the Stevens service robots are already at deployment, while the humanoid phase sits back at the training and data stage.

YYForce argues Singapore suits robotics adoption because service operators are trying to raise productivity, manage manpower constraints and keep service quality consistent. It points to a physical AI testbed at Punggol Digital District, announced in May 2026 by the Infocomm Media Development Authority, JTC Corporation and the Singapore Institute of Technology and planned for later in 2026, with applications including delivery, cleaning and security patrolling, as wider market context.

The two forecasts cover different, possibly overlapping categories, and YYForce says they shouldn't be added together. They are industry context rather than estimates of Singapore demand or YYForce's own addressable revenue. The Stevens deal follows a similar order: the confirmed first phase uses specialized service robots, while humanoid work sits in a separate second phase.

That order matches the company's stated plan for 2026 to 2031: evaluate specialized service robots for defined commercial tasks while developing humanoid applications through training and testing. The service robots are the near-term business. The humanoids are the bet.

When the projection meets the invoice

The next marker is the evaluation the partners have already promised: actual costs, productivity and guest-service outcomes compared with the 10% model.

Operators following the results have two more yardsticks to hold them against. Langkawi is tracking task completion, reliability, staff adoption and guest feedback, and YYForce's own deployment criteria add safety and required human supervision. Supervision is the one to watch for a hotel finance team, because staff time saved on delivery trips only counts if it isn't spent minding robots instead.

The partners say accompanying factsheets set out the deployment scope, the AI research programme and the longer-term strategy, which Fu tied to the company's commitment to "Forging Forward Together." YYForce's 2030 Vision sets the long-range target of an integrated workforce ecosystem. The hotels' evaluation of the service robots comes first.