Sometime in the next eighteen months, a humanoid robot is going to walk into a hotel lobby and try to help someone carry their luggage. It will not be the first time this has happened in a controlled demo or a press event. But it will be the first time it happens routinely, at scale, as part of an actual hospitality operation. And when it does, someone is going to have to answer a question that almost nobody in the robotics industry has seriously prepared for: what is that robot wearing?
It sounds trivial. It is not.
Where Things Stand Right Now
The humanoid robot sector in early 2026 is in a strange in-between phase. The machines exist. They work, to varying degrees. But they are almost entirely confined to environments where nobody except other workers sees them.
Tesla's Optimus units are operating in Tesla's own manufacturing facilities in Fremont and at the Austin Gigafactory, performing parts sorting, battery cell handling, and limited assembly tasks. The robots work behind barriers, in controlled environments, alongside human workers who have been briefed on their capabilities and limitations. Nobody cares what they look like. They are tools on a factory floor.
Figure's robots have been deployed at BMW's Spartanburg plant in South Carolina since late 2024, handling bin-picking and parts transfer tasks. Again, industrial settings. The robots wear nothing. Their appearance is purely functional: exposed actuators, visible wiring harnesses, the raw engineering on display. In a factory, that is fine. In a factory, it might even be preferable, because it signals "machine" clearly and reduces the chance of workers making false assumptions about the robot's capabilities.
Unitree's H1 and G1 platforms have made appearances at trade shows, corporate events, and tech demos across China, Japan, and increasingly in Europe and the United States. These are closer to public-facing scenarios, but they are still essentially performances. The robot shows up, does a choreographed routine, impresses people, and goes back in the case. The visual presentation at these events tends toward spectacle: sometimes the robot is bare, sometimes it is dressed in something eye-catching, but the choice is made for a camera, not for a Tuesday morning at work.
Agility Robotics' Digit units are in pilot deployments at Amazon and GXO Logistics warehouses. 1X Technologies' NEO is in early testing for residential applications. Sanctuary AI's Phoenix is in retail pilot programs in Canada. Across the board, the pattern is the same: controlled environments, limited public exposure, and almost no thought given to sustained visual presentation.
This is about to change.
The Transition Nobody Is Ready For
The trajectory is obvious to anyone watching the industry. Humanoid robots are moving from back-of-house to front-of-house. From factories to lobbies. From warehouses to showrooms. From labs to living rooms.
Hilton has been publicly discussing robot concierge programs since its early work with IBM Watson and the Connie prototype. That was a stationary kiosk robot. The next iteration will walk. Marriott International has filed patents related to autonomous room service delivery using bipedal platforms. Several major airlines have announced partnerships with robotics companies for terminal navigation assistance, though none have deployed humanoid units yet.
In corporate settings, the push is even more concrete. Japanese firms have been deploying humanoid reception robots for years (SoftBank's Pepper, now largely retired, was the pioneer), but the next wave involves robots that move through the building rather than standing at a desk. They deliver documents, guide visitors, carry equipment. They share hallways and elevators with employees and clients.
The residential market is the furthest out but the most consequential. 1X Technologies and Tesla have both signaled that home deployment is an explicit goal for their humanoid platforms. When a robot lives in someone's house, visual presentation is not a nice-to-have. It is a core product requirement. Nobody wants an exposed-actuator industrial machine standing in their kitchen.
Each of these transitions, factory to hotel, warehouse to office, lab to home, forces the same question: how should this robot look when non-specialist humans interact with it daily?
Why Appearance Is Not a Cosmetic Problem
There is a tendency among robotics engineers to dismiss visual presentation as a marketing concern. Something for the brand team to worry about after the real engineering is done. This is a mistake, and the research literature is unambiguous about why.
Human responses to robots are shaped heavily by visual first impressions, and those impressions form in milliseconds. A 2019 study by Natalia Calvo-Barajas and colleagues at Uppsala University found that people made stable judgments about a robot's competence, warmth, and trustworthiness within the first 500 milliseconds of visual exposure. Those snap judgments predicted interaction quality over extended periods. First impressions with machines, it turns out, work a lot like first impressions with people.
Christoph Bartneck's research group at the University of Canterbury has published extensively on the role of robot appearance in human-robot interaction. Their findings consistently show that surface-level visual cues, color, texture, clothing, perceived "grooming", affect not just how people feel about robots but how effectively they cooperate with them. A robot that looks trustworthy gets better compliance with its requests. A robot that looks competent gets more patience when it makes errors.
The implications for deployment are practical, not academic. A hotel robot that guests find unsettling will generate complaints. A corporate office robot that employees find off-putting will be underused. A home robot that family members find cold or threatening will be returned. In each case, the robot's capabilities might be excellent. The failure is in the presentation layer.
First impressions with machines work a lot like first impressions with people. You get about half a second.
There is also the uncanny valley problem, which is well-documented but still poorly managed in practice. Masahiro Mori's original 1970 observation, that humanoid entities become unsettling as they approach but fail to reach full human likeness, remains a real design constraint. Clothing is one of the most effective tools for managing it. Covering a robot's mechanical body while leaving its face and hands visible lets designers control exactly how much machine is on display. Too much exposed mechanism and people recoil. Too much attempted human-likeness and people feel deceived. Clothing provides a middle ground: a familiar social signal wrapped around an unfamiliar body.
Who Is Actually Building for This?
Given the obvious trajectory, you might expect a thriving industry of robot garment manufacturers gearing up for the deployment wave. The reality is thinner than that.
The industrial robot cover market is mature. Companies like ABB Robotics and KUKA have supplied flexible protective covers for articulated arms for decades. These are functional covers, typically made from durable technical fabrics, designed to protect against welding spatter, paint overspray, or particulate contamination. They are not designed for human-facing aesthetics. They are tarps with zippers.
On the humanoid side, the market is genuinely nascent. A handful of operations are building garments specifically for humanoid robot platforms, but the list is short.
Maison Roboto, a Paris-based couture house, has been constructing garments for humanoid platforms since 2024 and currently supports seven different robot models. Their work sits at the luxury end: bespoke pieces designed for specific platforms, with attention to sensor clearance, joint articulation, and thermal management. It is couture-level construction applied to non-human bodies. (L'Officiel Monaco profiled their approach earlier this year.) AVDI, a more recently launched label, is targeting the deployment tier with ready-to-wear robot garments designed for operators who need to outfit multiple units quickly and consistently.
Beyond these dedicated operations, there are scattered efforts. Several fashion schools, including Central Saint Martins and Parsons, have run student projects on robot garments. A few cosplay and costume fabricators have built one-off pieces for specific robots. Some robotics companies have internal design teams that produce basic covers or shells. But there is nothing resembling the kind of supply chain that large-scale deployment will require.
The contrast with the human uniform industry is stark. Companies like Cintas, Aramark, and UniFirst supply millions of garments annually to corporate clients, with full services for design, manufacturing, distribution, laundering, and replacement. The robot garment equivalent of this infrastructure does not exist. Not yet.
The Goldman Sachs Number and What It Actually Means
Goldman Sachs' widely cited projection estimates the humanoid robot market will reach $38 billion by 2035, with a meaningful share of that coming from service, hospitality, and residential applications. Other estimates are even higher. The numbers get thrown around frequently in investor presentations and press releases, usually to convey the message that this is a large and growing opportunity.
What those numbers do not capture is the secondary markets that deployment at that scale would create. If there are hundreds of thousands of humanoid robots working in hotels, offices, retail environments, and homes by 2035, every single one of them will need some form of visual presentation solution. Many will need multiple outfits for different shifts, seasons, or contexts. All of them will need replacements as garments wear out.
Run the math with conservative assumptions. Assume 500,000 humanoid robots in customer-facing roles by 2035 (well below what Goldman's projection implies). Assume each needs an average of three garment sets at any given time (a minimum for rotation and laundering). Assume each garment set costs $200 at the deployment tier (far below couture pricing, but above fast fashion). That is $300 million in initial outfitting alone, before you account for replacement cycles, seasonal updates, or premium tiers.
The point is not the specific number. The point is that the clothing question is not peripheral to the deployment story. It is embedded in it. Every robot that moves from a factory to a public space creates demand for visual presentation solutions. The $38 billion robot market implies a garment market that, while much smaller in absolute terms, will be an essential enabling layer.
What Deployment Operators Actually Need to Think About
For the operations managers, fleet supervisors, and hospitality directors who will actually be making these decisions, the clothing question breaks down into several concrete concerns.
Sensor Compatibility
Every current humanoid platform bristles with sensors: cameras, LiDAR, depth sensors, microphones, force-torque sensors, IMUs, and proximity detectors. Any garment that covers the robot's body must not occlude, interfere with, or degrade the performance of these sensors. This is a harder constraint than it sounds. A camera that works fine through clear air may lose significant resolution looking through a layer of fabric, even loosely draped fabric. LiDAR can be scattered or attenuated by textile fibers. Microphone sensitivity drops when covered by material with certain acoustic properties.
The practical solution is precision engineering: garments designed with sensor-specific cutouts, transparent windows, or acoustically transparent panels at exactly the right locations for each platform. This requires detailed knowledge of each robot's sensor layout, and it means that a garment designed for one platform generally will not work on another without modification.
Thermal Management
Humanoid robots generate substantial heat from motors, computing hardware, and battery systems. Most current platforms are designed to dissipate that heat through their outer shell. Covering that shell with fabric changes the thermal equation. A poorly designed garment can trap heat, causing the robot to throttle its performance or shut down entirely.
Good robot garment design accounts for this with breathable fabrics over heat-generating areas, mesh panels aligned with ventilation ports, and materials that do not insulate where insulation is not wanted. Some designers are experimenting with phase-change materials in garment linings that absorb excess heat and release it gradually, effectively extending the robot's operating envelope rather than constraining it.
Range of Motion and Joint Articulation
A humanoid robot's value comes from its ability to move like a human. Any garment that restricts that movement reduces the robot's utility. This is particularly challenging at the shoulders, elbows, hips, and knees, where the range of motion is large and the mechanical joint geometry does not match human anatomy.
Human tailoring has thousands of years of pattern-making knowledge for accommodating human joints. Robot joints are different. They rotate on different axes, through different ranges, with different geometries. The dart placements, gusset shapes, and stretch panel locations that work for a human shoulder do not necessarily work for a Unitree G1 shoulder or a Tesla Optimus shoulder. Each platform requires its own pattern engineering.
Durability and Maintenance
A robot in a commercial deployment might operate sixteen hours a day, seven days a week. Its garments will be subjected to far more wear than typical human clothing. They need to withstand repeated mechanical stress at the joints, frequent cleaning (some environments require daily sanitization), and the general abrasion of constant use.
They also need to be easy to put on and take off. A garment that requires thirty minutes and two technicians to install is a garment that will not get changed when it should. Quick-release fastening systems, modular construction, and designs that account for the specific geometry of each robot's limbs and joints are essential for practical deployment.
Brand Consistency
For corporate deployments, the robot's appearance needs to align with the operator's brand identity. This goes beyond slapping a logo on a polo shirt. It means the robot's entire visual presentation, colors, fabrics, style, fit, level of formality, needs to feel coherent with the brand's existing visual language. A luxury hotel's robot should look like it belongs in a luxury hotel. A tech company's office robot should look like it belongs in a tech company's office.
This is a design challenge that sits at the intersection of brand strategy, fashion design, and robotics engineering. It is not a skillset that exists in abundance.
The Gap Between Here and There
The honest assessment of where the robot garment industry stands in April 2026 is: early. Very early. The machines are almost ready for public deployment. The clothing is not.
There is no standardized sizing system for robot garments. Each platform has its own dimensions, joint locations, sensor placements, and thermal profiles. A garment designed for one robot does not fit another, and there is no equivalent of the S/M/L/XL system that allows approximate fit across a range of body types. Every new platform requires new patterns.
There is no established supply chain for robot garment materials. The technical textiles needed, sensor-transparent fabrics, thermally engineered composites, high-durability stretch panels, exist individually but are not available as a coordinated toolkit. Robot garment designers are often sourcing materials from aerospace, medical textile, and performance sportswear suppliers, combining materials never intended to be used together.
There is no regulatory framework specifically addressing robot clothing. Safety standards for robots exist (ISO 13482 for personal care robots, ISO 10218 for industrial robots), but they do not address garments in any detail. Questions about flammability standards, visibility requirements, sensor occlusion limits, and maintenance hygiene standards for robot garments are currently unanswered at the regulatory level.
There is no workforce trained in this discipline. Robot garment design requires a combination of fashion design, textile engineering, robotics knowledge, and an understanding of human-robot interaction psychology. Almost nobody has all of these skills. The people who are good at this have mostly taught themselves through trial and error.
And there is limited awareness among robotics companies themselves that this is a problem they need to solve. Most humanoid robot manufacturers are focused, understandably, on locomotion, manipulation, perception, and AI. The question of what the robot wears when it meets a hotel guest is not on the critical path for any engineering team. It will only become urgent when the first large-scale deployments encounter the predictable public reaction to bare mechanical bodies in social spaces.
What Comes Next
The deployment wave is coming. Goldman Sachs, Morgan Stanley, ARK Invest, and every major robotics company's own projections agree on the direction, even if they disagree on the timeline. Humanoid robots will move into public spaces. They will work alongside people. They will interact with customers, guests, patients, and family members.
When that happens, the clothing question will stop being speculative and start being operational. Operations managers will need garment solutions that work with their specific robot platforms, fit their brand requirements, survive commercial-grade use, and ship in volume. They will need them on a timeline measured in weeks, not months.
The companies and designers who are building that capability now, figuring out the materials science, the pattern engineering, the sensor integration, the supply chain logistics, will have a significant head start. Everyone else will be scrambling.
The deployment question is not really about whether robots will enter public spaces. That is settled. The question is whether anyone will be ready to dress them when they do.
For background on why clothing matters for robots at all, see Why Robots Wear Clothes. For the broader commercial landscape, see our industry overview. For platform-specific guidance, the humanoid robot fashion guide covers the major platforms in detail.