Tag Archives: Robots

Insta360’s foldable drone patent targets DJI Neo 2 at sub-250g

Posted on by

Insta360 built its reputation on cameras that see in every direction at once. It’s literally in their name. The X series made 360-degree capture accessible; the Luna Ultra brought Leica optics to a handheld gimbal. Flying something, though, is a different category of challenge, and the patent images that appeared in July 2026 show the company working through what a foldable drone looks like with Insta360 DNA. The illustrated design at the bottom shows folding arms, a ring-shaped propeller guard, and a dedicated locking mechanism, a form factor that prioritizes the moment of packing away just as much as the moment of takeoff.

DJI’s Neo 2 has the casual drone segment largely to itself in the sub-250g category, with a rigid one-piece frame and a 19-minute flight time that have made it a reliable default for creators who want aerial footage without the complexity of a serious drone setup. Insta360 entering that space would break what has effectively been a two-player market, with Zero Zero Robotics’ HoverAir as the only real alternative. The folding design alone gives Insta360 a structural argument against the Neo 2; a potential 360-degree camera, if it materializes, would give it a creative one. What a wonderful year for foldables, eh? We’re expecting a folding iPhone in September, and Insta360 might give us a folding drone too!

This isn’t just some forum rumor, either. The patent images (from patent file CN223865117U), which started making the rounds on July 16th and were highlighted by outlets like Tech Court, are pretty specific. They show a drone body with the “Insta360” logo stamped on it, with fully folding arms that tuck into the chassis and a distinctive ring-shaped guard protecting the propellers. There’s even a dedicated locking mechanism detailed, which tells you just how seriously they’re taking the whole “fold-and-go” concept. This is a design that screams it was made to be stuffed into a pocket, not carefully placed in a bespoke case.

But here’s the thing: this isn’t Insta360’s first aerial rodeo. Before they decided to put their own name on a drone, they sent in a scout team under the sub-brand Antigravity. The A1 drone they launched was, frankly, a bit of a monster: a sub-249g drone with a wild dual-lens setup that shot 8K 360-degree video and made the drone itself invisible in the footage. It was a spec-sheet beast and a clear shot across DJI’s bow, proving Insta360 had the engineering chops to compete in the sky. It was their way of testing the waters.

Now, it looks like they’re done testing. This new patent is the second, more direct shot. And here’s the kicker: for a company obsessed with 360-degree video, the patent images don’t show a fisheye lens. At all. It’s a huge question mark hanging over the whole project. Are they hiding the tech, or are they planning to launch a drone that, for the first time, doesn’t have their signature feature? If they’re aiming for a rumored sub-$200 price point, leaving the 360 camera on the cutting room floor might be the only way to get there. It’s a move that would be both shocking and, in a weird way, perfectly logical.

You also can’t forget that this is all happening while the two companies are actively suing each other in China. This isn’t friendly competition; it’s a full-blown hardware rivalry. Insta360 came for DJI’s gimbals, tested the drone market with a sub-brand, and is now preparing a direct assault on DJI’s most accessible drone. With a release window expected somewhere between late 2026 and early 2027, we’re not just watching a new product being born. We’re watching the next chapter in a tech soap opera, and it’s about to get very, very interesting.

The post Insta360’s foldable drone patent targets DJI Neo 2 at sub-250g first appeared on Yanko Design.

This One-Wheel robot could patrol your property while you stay away

Posted on by

Most security robots are designed like small autonomous cars, with a bulky body and enough room for all the fancy tech. 1ROLLO, however, adapts a radically different approach. Developed by Estonian startup Rollo Robotics, this autonomous security robot balances on a single large wheel, turning an unusual piece of robotics into a practical patrol machine for large properties.

The 1ROLLO is built around a vertically oriented spinning flywheel housed inside its main wheel. Gyroscopic forces help the bot remain upright while moving, while the internal mechanism allows it to lean and change direction. The result is a remarkably narrow robot that can move independently through spaces that might be less accessible to larger wheeled security platforms.

Designer: Rollo Robotics

Despite its unconventional appearance, the 1ROLLO surveillance robot is designed for serious security work. Its autonomous navigation system combines 360-degree computer vision with RTK GNSS positioning, allowing the robot to follow mapped patrol routes while navigating around obstacles. It can detect movement and objects, record video, recognize license plates, and monitor for potentially dangerous gases, including carbon monoxide. The idea is to shift routine security patrols from people to machines without completely removing human oversight. This is estimated to save up to 80 percent of the cost that would otherwise be incurred using manual methods.

When the robot detects something unusual, it can alert a remote operator over 4G, 5G, or Wi-Fi. The operator can then access a live video feed, speak to people near the robot using its microphone and speaker, or dispatch a human guard if the situation requires intervention. A cloud-based management platform is intended to make the system scalable, allowing one operator to monitor multiple robots simultaneously. That could make the 1ROLLO particularly useful for expansive sites where continuously patrolling every corner would require significant manpower.

The robot measures approximately 60 x 60 x 140 centimeters and weighs around 35 kilograms according to the company’s specifications, while more recent reporting has described a 45-kilogram version with a higher top speed of up to 30 km/h. The company has also stated a battery runtime of up to eight hours, after which the robot can autonomously return to a charging station. Its IP65-rated enclosure protects against dust and water jets, supporting outdoor operation.

Rollo Robotics sees potential applications across airports, data centers, industrial facilities, corporate campuses, manufacturing sites, hospitals, hotels, and critical infrastructure. The robot could also find a role in environments such as mining operations, construction sites, oil and gas facilities, and gated communities, where autonomous monitoring could complement existing security teams.

Rollo Robotics plans to offer the surveillance robot through a Robotics-as-a-Service model. Customers would pay for access to the system while receiving maintenance, software updates, support, and future hardware upgrades as part of the service. The 1ROLLO is still evolving, with the company working to improve its performance in challenging weather and busy environments. Commercial availability is expected in 2027, but its single-wheel architecture already makes it an intriguing alternative to the increasingly familiar fleet of four-wheeled security robots.

 

The post This One-Wheel robot could patrol your property while you stay away first appeared on Yanko Design.

SAM Might Be the Most Human Thing Samsung Has Finally Built

Posted on by

Most conversations about aging and technology eventually circle back to the same frustrating gap: the tools we build for older people rarely feel like they were designed with older people in mind. They’re functional, clinical, and often carry the visual language of a hospital supply catalog. SAM, a concept robotic mobility aid designed by David Webber for Samsung, feels like a direct response to that problem, and it’s the kind of design thinking I genuinely want more of.

At first glance, SAM looks like it belongs on a near-future film set. The rounded, cloud-white body, the articulating arms, the small expressive face-like display on top. But spend a little more time with it and the aesthetics start to make sense in a very deliberate way. Webber didn’t want SAM to read as medical equipment. He wanted it to feel like a companion, something you’d actually invite into your home and not hide in a corner when company comes over.

Designer: David Webber for Samsung

What SAM does differently from most mobility aids is where the real story lives. A standard walker or rollator solves one problem: moving from point A to point B. SAM looks at the fuller picture of what daily independence actually requires. Picking up a dropped object from the floor. Carrying groceries from the door to the kitchen. Helping someone rise from a chair safely. These are the small, invisible tasks that quietly determine whether an older adult can continue living alone or not, and they are almost entirely absent from the current conversation around mobility technology.

The concept is built around the idea of physical AI, the notion that artificial intelligence should operate in the physical world in a way that supports human capability rather than overriding it. SAM doesn’t take over. It doesn’t assume. The user stays in control, and the technology fills in the gaps, which is a harder design problem than it sounds. Building something that assists without patronizing, that helps without making a person feel like they’ve surrendered something, requires a kind of empathy that most tech doesn’t bother with.

I’ll be direct: the design of SAM is one of the more thoughtful things I’ve seen come out of a Samsung concept in a while. The color options, soft greys and muted sage greens, feel intentional. They’re domestic. They say “living room” rather than “rehab facility,” and that distinction matters enormously for user dignity and adoption. Nobody wants to use a tool that announces their limitations to every visitor who walks through the door. SAM seems to understand that.

The arms and end-effectors are worth a close look too. The gripper design is specific enough to handle real objects but rounded and soft-looking enough not to feel threatening. The integrated display sits at a natural eye level when someone stands and grips the handles, showing simple information like time, a to-do list, and device status. It’s thoughtful UX layered into an already thoughtful form.

Is SAM a product you can buy right now? No. It’s a concept, a vision piece, and it should be taken as one. But the best concept designs do something important: they shift where the conversation is pointed. Right now, too much of the assistive tech space is focused on reactive solutions, devices that help after a fall, after a diagnosis, after independence has already started to erode. SAM is proposing something more proactive. It’s asking what support looks like before the crisis, and that’s a question the industry desperately needs to sit with.

What I hope designers, healthcare companies, and tech brands take from SAM is not the specific form but the underlying philosophy. That independence isn’t just about mobility. That dignity is a design requirement, not a bonus feature. That the people who need the most thoughtful design often receive the least considered version of it. If SAM pushes even one conversation in that direction, David Webber has done something genuinely worth paying attention to.

The post SAM Might Be the Most Human Thing Samsung Has Finally Built first appeared on Yanko Design.

Weave’s $8,000 Isaac 1 Is the Home Robot That Dresses Like Furniture

Posted on by

Most robots designed for the home look like they escaped from a manufacturing plant. They’re shiny, angular, and cold, and the moment you put one next to your sofa, the visual mismatch is almost comical. So when I came across the Isaac 1 by Weave Robotics, the first thing I noticed wasn’t what it could do. It was how it looked: soft, compact, wrapped in swappable fabric shells in colors like Sage, Slate Blue, and Terracotta. It looked less like a machine and more like something you’d find at a furniture fair. That’s entirely by design, and it matters more than most people probably realize.

Weave Robotics built the Isaac 1 from the ground up in San Francisco, and their approach is refreshingly considered. Instead of the humanoid form that most robotics companies seem obsessed with, Isaac 1 runs on a wheeled base and collapses down to about three feet when it’s not working, which means it can genuinely disappear into your space. When it’s needed, it extends up to 5’9″ with an 80-inch vertical reach and a 38-inch horizontal reach, enough to handle a wide range of domestic tasks without needing to replicate the human body to do it.

Designer: Weave Robotics

And the tasks it handles are, frankly, the ones most of us quietly dread. Laundry is front and center. The Isaac 1 picks up dirty clothes, manages loaded hampers, folds, and puts things away. It also handles the daily reset of a home: making beds, straightening cushions, returning clutter to where it actually belongs. If you’ve ever come home to a messy house at the end of a long day and felt that specific kind of low-grade despair, you understand the appeal.

But here’s what gets me about the Isaac 1. The decision to wrap it in fabric rather than hard plastic isn’t just cosmetic. It signals something about how Weave Robotics thinks about the relationship between people and the technology they invite into their homes. A robot that reads as furniture rather than lab equipment is a robot that respects the space it’s entering. It acknowledges that a home is a personal, lived-in place, not a test environment. That philosophical underpinning makes Isaac 1 feel different from the robots that tech companies roll out at trade shows and then quietly retire two years later.

Privacy is also built into its physical design, not just its software policy. Isaac 1 has physical cues that make it visually clear when it’s active versus idle, and its cameras can be physically disabled. For a connected device that moves through your most private rooms, that kind of intentionality is worth noting. The teleoperation feature, where a human can step in remotely when the robot hits an edge case it can’t solve autonomously, is the detail that deserves the most scrutiny. Weave has been transparent about it, but it’s something you want to sit with before committing.

That commitment currently sits at $8,000, with pre-orders open and first deliveries expected in California in fall 2026, and wider US availability following in 2027. It’s not an impulse buy by any measure. The price positions it well above a novelty and firmly in the category of household infrastructure, which, to be fair, is exactly what Weave is pitching it as.

I don’t think the Isaac 1 is perfect or even fully proven yet. Homes are unpredictable environments and no amount of smart engineering fully accounts for the everyday chaos of real life. But as a statement of intent about what a domestic robot can and should be, it’s one of the more coherent things I’ve seen come out of this space. It’s not trying to look human. It’s not trying to impress you at a conference. It’s trying to fit into your home, quietly, practically, and without making a scene. That’s harder to achieve than it sounds, and the fact that Weave seems to understand that puts them ahead of most.

The post Weave’s $8,000 Isaac 1 Is the Home Robot That Dresses Like Furniture first appeared on Yanko Design.

DJI built a parachute into their drones to protect the hardware… and people’s heads

Posted on by

Drones fall. Sensors fail, signals drop, batteries die at inconvenient moments, and until now the consequences of that failure were left largely to chance. Obstacle avoidance technology solved one half of the safety equation, keeping drones from flying into buildings, power lines, or each other. But avoidance only works while the aircraft is still flying. Once a drone loses power or control, all that sensor smarts becomes irrelevant, and physics takes over. Even a small consumer grade drone weighing 8 or 9 ounces turns into a falling object with real force behind it when it drops from 200 or 300 feet, and DJI’s Matrice 400 is a far heavier machine flying far more demanding missions.

That is the gap the AP100 Parachute was built to close. It watches its own hardware in real time, running continuous checks so it knows the instant something looks off. Backup capacitors keep the system powered even if the drone itself loses electricity, which matters because a mid air failure rarely comes with a warning. Once triggered, deployment is instant, and the parachute slows descent to under 5 meters per second, turning what would be a crash into something closer to a controlled landing. Alarms then kick in after deployment, alerting anyone nearby and helping pilots track down exactly where their aircraft ended up.

Designer: DJI

The AP100 gives pilots three separate ways to pull the ripcord, so to speak. It can trigger itself automatically the moment it detects an anomaly or a geocaging breach, meaning the drone strays somewhere it should not be. Pilots who want manual control can swipe to deploy directly inside the DJI Pilot 2 app, no digging through menus required. There is also a remote option through FlightHub 2’s FTS page, useful for fleet operators managing multiple aircraft from a central hub. Layering automatic, manual, and remote deployment together means there is rarely a scenario where the parachute simply cannot be triggered in time.

None of this exists in a vacuum. Commercial drone missions flying over cities or beyond the pilot’s direct line of sight, what the industry calls BVLOS, have to satisfy strict safety categories like C5 and C6 before regulators sign off. Surveying a highway route or inspecting a pipeline that snakes through populated areas used to mean stacking on extra precautions just to get approval. A parachute rated to meet those standards effectively does the compliance heavy lifting on its own. That turns what used to be a drawn out approval process into something considerably more straightforward for operators trying to get missions off the ground, literally and legally.

The hardware itself is built for the grind of daily operations rather than sitting as a one time safety gimmick. Pilots can swap batteries without ever detaching the parachute, which sounds minor until you are the one doing it in the field between flights. It also carries the same IP55 rating as the Matrice 400 itself, so rain or dust does not take it out of commission. Full system self checks run automatically at startup, cross referencing communication links and hardware status before the drone even leaves the ground. These are the kinds of details that matter more to the operator running ten missions a week than to anyone reading a spec sheet once.

Here is the catch, though. The AP100 is built exclusively for the Matrice 400, DJI’s enterprise platform aimed at industrial inspection, surveying, and public safety work. Your Mavic will not be getting one, and neither will your FPV drone, at least not yet. That is not entirely surprising given the size and cost tradeoffs involved in strapping a parachute system to something meant to be light and nimble. But DJI has a track record of trickling enterprise safety features down into consumer lines once the engineering matures, and a company that just built a first party parachute for one drone rarely stops at just one.

Regulators keep pushing drones toward busier skies, over highways, over crowds, over exactly the kind of scenarios where a hardware failure stops being a minor inconvenience and starts being a genuine hazard. DJI building a parachute directly into its own ecosystem, rather than leaving it to third party add ons, suggests the company sees this as a standard feature rather than an optional extra. Whether that logic eventually reaches the Mavic or Air lineup remains to be seen, but the direction feels obvious. Falling should not be part of the risk calculus anymore, and DJI just made a very public bet on that idea.

The post DJI built a parachute into their drones to protect the hardware… and people’s heads first appeared on Yanko Design.

The Air Purifier That Rolls Toward Smoke Before It Fills the Room

Posted on by

Air purifiers have quietly become household staples, tucked into corners and set to auto, doing their work without much acknowledgment. The fundamental assumption behind every stationary model is that polluted air will eventually drift past the filter if you wait long enough. That’s not entirely wrong, but it means the unit is always reacting to something that’s already spread rather than catching it at the source.

Omnipure reframes the whole premise. Rather than sitting in one corner hoping the air comes to it, this autonomous robot concept maps the home, learns the occupant’s routines, and moves toward pollution before it has a chance to disperse. Catching cooking smoke at the stove rather than after it fills the room is the kind of meaningful difference that makes a purifier genuinely useful rather than marginally helpful.

Designer: Kyuhong Kim

The behavior is where the design makes its most interesting decisions. When someone walks through the front door, Omnipure rolls to the entryway and greets them while clearing entryway dust at the same time. Framing that as a welcome rather than a clinical alert is a deliberate choice. A robot charging toward you the moment you arrive home communicates something quite different from one that simply seems glad you’re back.

The kitchen scenario captures the same thinking. When cooking raises PM2.5 levels or volatile organic compounds spike from a hot pan, Omnipure navigates toward the source and announces its arrival with “Smells Good!” rather than an alarm. If TVOC levels climb too high, it suggests opening a window rather than filtering at maximum capacity, because the better response to a ventilation problem isn’t a better filter.

The ventilation behavior adds another dimension. Rather than treating indoor air as automatically preferable to outdoor air, Omnipure monitors both and suggests airing out when outdoor conditions are favorable. When a window is opened, it positions itself in the airflow to filter incoming air before it circulates. The messages it displays at these moments frame its function as a conversation rather than a status readout.

The face display communicates air quality through four states tied to PM2.5 levels. Clean air below 15 μg/m³ reads as a blue glow with relaxed eyes, acceptable conditions around 35 μg/m³ show green, poor air at 75 μg/m³ appears amber, and hazardous readings above 76 μg/m³ trigger a red glow with X-shaped eyes. A glance across the room is enough to know what you’re breathing without unlocking a phone.

Internally, Omnipure runs on LiDAR and time-of-flight sensors for navigation, a clipping camera, dual temperature and humidity sensors, and a centrifugal ventilation fan drawing air through a large replaceable filter housed in the lower body. The filter slides out from the side for maintenance, which turns what’s usually a buried, tool-required task into something that takes about as long as changing an ink cartridge.

Omnipure is a concept rather than a product on shelves. But the problem it identifies is real: stationary purifiers can only clean the air that eventually reaches them. A device that maps a home, learns activity patterns, and positions itself where pollution is forming solves a fundamentally different problem from one that simply runs on auto in a corner, hoping the air comes to it.

The post The Air Purifier That Rolls Toward Smoke Before It Fills the Room first appeared on Yanko Design.

Harvard Students Built a Lamp That Taught a Robot to Make Art

Posted on by

There are moments in design when your first thought isn’t “how does it work” but “how does it feel.” Hikarigami is one of those objects. Created by four Harvard Graduate School of Design students, Luke Fiorante, Joseph Fujinami, Annie Xing, and Chi Zhang, it’s a luminaire that looks, at first glance, like an intricate metal tower carved by a very patient and very precise hand. But the story behind it is far more interesting than the object alone, and the object itself is already extraordinary.

The name is a portmanteau of two Japanese words: hikari, meaning light, and kirigami, the traditional art of cutting and folding a single sheet of material into three-dimensional form. The fusion isn’t just poetic; it’s structural. The team took that ancient paper-craft logic and translated it into sheet aluminum, using an industrial robotic arm to press, deform, and expand the metal into a lattice of thousands of individual cells. Each cell is unique. Each one catches light differently. And every fold was made without a mold.

Designers: Luke Fiorante, Joseph Fujinami, Annie Xing, Chi Zhang

That last part is the quiet revolution tucked inside this project. Traditional metal forming depends on expensive, custom-made dies and molds, static tools that lock a design into one fixed shape and require significant capital before a single piece can be produced. Hikarigami throws that model out entirely. The team developed a custom script called Machina, which programs the robot to adjust its actuation depth for every single cell in the lattice. Because the cells vary in size across the panel, a uniform press would cause the metal to fail. The script scales the force to each cell’s geometry, working right up to the material’s yield limit without crossing it. It’s precision at a scale that a human hand simply couldn’t replicate.

But precision isn’t the point. Craft is. The team was genuinely asking whether robotic automation could learn the intuitions of handmaking: the responsiveness to material, the calibration to context, the sense that a gesture should change depending on where you are in the process. Most industrial fabrication doesn’t care about any of that. Hikarigami does. And to me, that’s the more interesting question.

The result is a tower made of six identical aluminum panels with no fasteners, no adhesives. The panels interlock through tabs built directly into the geometry, which means the whole assembly is a single material: pure aluminum, fully recyclable, and designed to age gracefully. The team noted that over time, the aluminum will develop a patina shaped by its environment. The lamp doesn’t resist change. It participates in it.

When you turn it on, the experience shifts completely. During the day, Hikarigami reads as architectural, almost industrial, the kind of object that would feel at home in a gallery or a thoughtfully designed living space. At night, with the central LED filament lit, the aluminum skin opens up. The thousands of robotically formed apertures cast overlapping pools of shadow and caustic highlights across every nearby surface. The room becomes part of the lamp. The lamp becomes the room.

It’s worth noting that this is a student project, and I find that genuinely refreshing. It won in the sustainability category at one of design’s most respected annual competitions, and it also picked up a notable distinction in furniture and lighting. For work produced inside an academic program, the level of technical and conceptual rigor here is striking. Beyond the accolades, Hikarigami feels significant because it takes a position. It argues, quietly but clearly, that sustainable design doesn’t have to mean sacrificing beauty or experimentation. That mold-free fabrication can be expressive. That a robot can be a collaborator, not a replacement.

The team divided their roles with intention: Xing led computational design and fabrication, Fiorante handled robotic toolpath programming, Fujinami managed structural assembly, and Zhang contributed to the robotic fabrication process. Together they didn’t just design a lamp. They designed a process, and the lamp emerged from it. That’s the kind of thinking that tends to outlast the object itself.

The post Harvard Students Built a Lamp That Taught a Robot to Make Art first appeared on Yanko Design.

XiaoBu Wants to Be the Most Personal Thing on Your Desk

Posted on by

We’ve reached a point where our desks have become a form of personal expression. The plant you chose, the pen holder you 3D-printed, the specific coffee mug that only leaves when it’s being washed – these things say something about who you are. XiaoBu, a desktop companion robot concept by Handsome Chen, is betting that your next statement piece might just say “hello” back.

XiaoBu is not a smart speaker with a face. It’s not a Tamagotchi with an engineering degree. It’s a desktop robot that pulls together three things that usually live separately: audio function, emotional interaction, and replaceable fabric skins. The combination is what makes it feel genuinely interesting rather than like another product designed to solve a problem no one actually had. Sit with the concept for a moment, though, and the sum starts to feel quite different from its parts.

Designer: Handsome Chen

The fabric skins are the detail that stopped me mid-scroll. Most companion robots lean hard into hard plastic shells – which makes sense from a durability standpoint, but creates a certain coldness that works against the whole “companion” pitch. XiaoBu takes the opposite direction. By making the outer layer soft and swappable, it invites you to treat it less like a gadget and more like a small, quietly present roommate you can dress according to your mood. That shift in material logic carries a real design philosophy behind it: the idea that emotional connection often starts with touch, texture, and familiarity. It’s the difference between something you pick up because you need it and something you keep because it just feels right having it nearby.

The trend toward desktop companion robots has been building steadily for some time. Products like EMO and Eilik have already carved out a niche for people who want their desk to feel a little more alive – but those largely stick to the personality-through-expression model, where a tiny LED screen becomes the entire emotional vocabulary of the machine. XiaoBu’s approach is quieter and, I’d argue, more considered. It’s not performing for you. It’s designed to live alongside you, and the form reflects that distinction deliberately.

Handsome Chen, the Xi’an-based designer behind the concept, describes XiaoBu as breaking the “static limitations” of traditional desktop objects. It’s a deceptively simple framing, but it gets at something real. Most things on your desk are passive. They sit. They wait. They do nothing unless you reach for them. A companion robot, by design, has to earn its place differently – not just by being useful, but by being present in a way that feels intentional rather than intrusive.

I’ll be upfront: I’m a little skeptical of the word “companion” being attached to consumer tech. It’s an easy word to deploy and a hard promise to keep. But XiaoBu’s design, at least conceptually, seems to understand that emotional resonance can’t be engineered through features alone. It has to be built into the form, the texture, the way the object occupies space. The replaceable skins aren’t just a customization feature – they’re an acknowledgment that the relationship between a person and their space is fluid and always evolving, and the things they keep close should be allowed to evolve too.

What makes XiaoBu feel genuinely fresh in this growing category is that it doesn’t try to impress you with complexity. It doesn’t seem to be chasing a checklist of AI capabilities or competing on specs. The design conversation starts with softness, with audio warmth, with the kind of small, considered decisions that make an object feel like it belongs to you rather than to a product category.

Whether XiaoBu ever moves past concept into production is a separate question, and one I genuinely hope gets answered. Right now it lives on Behance, accumulating thousands of appreciations from people who clearly recognize the same thing I did: that sometimes the most interesting design isn’t the one with the most features. It’s the one that makes you feel like someone actually thought about what it means to share space with you.

The post XiaoBu Wants to Be the Most Personal Thing on Your Desk first appeared on Yanko Design.

Forget Humanoids: Eno Might Be the Robot We Actually Need

Posted on by

The robot race has been moving in one direction for a while now: two legs, a head, and a shape that clearly spent considerable time studying what a person looks like. It’s a logical instinct. Factories, hospitals, and homes were all designed with human proportions in mind, so a robot built like a human should, theoretically, slot right in. But a growing argument exists that this whole approach might be overthought, and Genesis AI’s new Eno robot is making that case louder than most.

Unveiled this week, Eno is the debut robot from Genesis AI, a San Carlos-based startup that quietly raised $105 million in seed funding and spent that time building something genuinely different. Eno doesn’t walk. It rolls. Its body is a minimalist, articulated tower rising from a wheeled base, with no face and no head in sight. The form adjusts its height and reach as needed, folds down for storage, and carries a pair of proprietary dexterous hands designed to move with the kind of precision and range that human hands are capable of. The result looks more like a sleek industrial sculpture than any robot you’ve seen at a tech keynote.

Designer: Genesis AI

That feels completely intentional. Genesis AI’s head of design, Daniel Hundt, has said that Eno was built by asking a single question first: what does the robot actually need to be? The answer stripped away everything decorative and kept everything functional. No face, because a face isn’t a prerequisite for doing work well. No legs, because legs add cost, complexity, and a surprising number of ways for something to go wrong. What remained was a form built around capability, not aesthetics trying to pass as capability. That’s a meaningful distinction in an industry that sometimes confuses the two.

Eno runs on GENE, Genesis AI’s proprietary robotics-native AI foundation model, and the two were developed together as a single integrated system. This matters more than it might initially seem. A lot of robots in this space are essentially off-the-shelf AI bolted onto hardware that wasn’t designed with it in mind. GENE and Eno were built to complement each other, which means the robot can reason through multi-step tasks, adapt when conditions change, and plan across long time horizons rather than just responding to simple, pre-defined commands. That kind of sustained, adaptive thinking is what separates a useful robot from an expensive demo reel.

For those who want a deeper look at what’s happening under the hood, Genesis AI is offering an optional screen version of Eno featuring a cognitive interface that displays, in real time, what the robot is thinking and processing. It’s an unusual transparency move for a robotics company, and a genuinely smart one. Trust in AI systems tends to erode when people feel like they’re watching a black box make decisions. Showing the work, quite literally, is one way to build confidence in environments where precision matters, like hospitals, labs, or busy production floors.

Deployment is set to begin with industrial customers by the end of 2026, starting with manufacturing, logistics, and laboratory settings before moving into service industries like hotels and hospitals, with consumer and home use following down the line. That rollout sequence makes sense. Controlled industrial environments offer a much cleaner test case for a robot learning the real world than someone’s living room does, and it gives Genesis AI the chance to refine Eno where the stakes of a miscalculation are measured in efficiency rather than anything more personal.

Whether Eno ends up being the robot that finally makes good on the promise of general-purpose robotics remains to be seen. The industry has announced breakthroughs before and delivered timelines that stretched well past the original projections. But Eno feels different in at least one significant way: it isn’t trying to win you over with its looks. It’s making a functional argument, and that alone puts it in a category of its own right now. Sometimes the smartest design choice is knowing exactly what not to include.

The post Forget Humanoids: Eno Might Be the Robot We Actually Need first appeared on Yanko Design.

KEENON Humanoid Pours Drinks at GCS 2026, 100,000 Others Run Hotels

Posted on by

The robotics industry has a curious reputation problem. The machines getting the most attention, walking bipeds that do backflips, aren’t the ones driving real business value. By 2030, professional service robots are projected to account for $90 billion of a $161 billion global market, growing at 24.6% annually. That makes them the fastest-growing segment in robotics, which is a fact that barely makes the news.

KEENON Robotics has known this for a while. Founded in 2010, the Shanghai-based company didn’t wait for the hype cycle; it built the service delivery robot category from the ground up. Today, with over 100,000 units shipped across more than 60 countries, KEENON holds the number-one global market share in commercial service robots for the third consecutive year, according to IDC 2025.

Designer: KEENON Robotics

At Global Connect Show 2026, the star of KEENON’s booth was the XMAN-R1, a wheeled humanoid robot that makes popcorn, pours drinks, and hands out snacks. It’s the kind of demo that stops foot traffic, and it’s meant to. Underneath the theatrics, though, is a robot packing 275 TOPS of AI processing power, dual 7-DoF arms, and precision dexterous hands built for human-level manipulation.

What the demo doesn’t show is how much work went into teaching a robot to grab a cup. KEENON estimates that a single action of that kind requires at least 1,000 data points. A full coffee-making sequence demands over 20,000. That gap between what looks effortless and what it actually costs computationally is one of the clearest explanations of where physical AI sits right now.

KEENON is remarkably candid about this. Their own assessment puts the current “mind age” of humanoid robots at roughly three years old, which, if you think about it, explains a lot about why they move so deliberately. True general-purpose humanoid deployment is still at least five years out by their estimation. The “Model T” framing they use is apt; these are early machines, not finished ones.

That candor is also what makes KEENON’s established product line feel more credible. While the XMAN-R1 gets the headlines, KEENON’s delivery and cleaning robots have been running inside hotels, restaurants, hospitals, airports, and casinos across more than 600 cities globally. Their DINERBOT T10 can carry up to 40 kg, fits through a 59 cm passage, and operates for up to eight hours on a single charge.

A good example of what that looks like in practice is Shangri-La Hong Kong, where KEENON runs six different robot types across eight units within a single hotel. Delivery bots handle contactless room service; cleaning robots run scheduled cycles through lobbies and corridors; logistics carriers shuttle linens and supplies behind the scenes. None of this required the hotel to restructure its operations around the robots.

Part of why that integration works comes down to a deliberate design choice. KEENON chose not to make its service robots look human. The compact, rounded bodies, soft voices, and animated screen faces are intentional, because how a robot looks determines whether people trust it. Western audiences carry Terminator-shaped anxieties; Asian audiences grew up with characters like Doraemon. The design has to work for both.

The XMAN-R1 at the Global Connect Show is KEENON’s way of signaling where the product line is heading. Alongside it, the company also offers the bipedal XMAN-F1, a full-body humanoid with 43 degrees of freedom that takes the same task-driven approach. Both run on KEENON’s KOM 2.0 platform and are designed to work alongside existing robots like the DINERBOT T10 rather than replace them.

What KEENON actually has over the wave of humanoid startups entering the market is something harder to replicate than a robot body. With 100,000 units running across 70+ countries, the company has been accumulating proprietary operational data at a scale most competitors haven’t even started to approach. Every delivery, every corridor, every cup poured feeds back into that picture, one data point at a time.

The post KEENON Humanoid Pours Drinks at GCS 2026, 100,000 Others Run Hotels first appeared on Yanko Design.