Household Robots in 2026: What Is Real, What Is Hype, and What Comes Next

For decades, people have imagined a general-purpose household robot that can cook dinner, fold laundry, wash dishes, clean every room, care for an older family member, and hold a natural conversation.

In 2026, that complete vision is still mostly a research and pre-market story.

But household robotics itself is no longer science fiction. Tens of millions of consumer service robots are sold each year. Robotic vacuums are common household appliances. Robotic lawn mowers, pool cleaners, window cleaners, companion robots, educational robots, security robots, and other specialized machines are commercially available across the world.

The reality is therefore more interesting than either extreme:

• Household robots are not merely hype.
• General-purpose robotic housekeepers are not yet mature.
• Specialized household robots are already a significant global market.
• Research progress in manipulation, perception, AI, simulation, and humanoid robotics is accelerating.
• The boundary between a smart appliance and a capable household robot is beginning to shift.

The challenge for consumers, researchers, engineers, investors, journalists, and manufacturers is that information about this field is scattered across product websites, research papers, company announcements, videos, datasets, code repositories, crowdfunding campaigns, and news reports.

That is the problem Max Robotics is being built to solve.

Max Robotics currently provides natural-language discovery across approximately 500,000 robotics and AI papers, using embeddings, vector search, and structured data. Instead of searching only by title, author, or exact keyword, users can ask questions in ordinary language and discover semantically relevant research.

The next stage is to connect that research layer to the real-world household-robot market: commercial products, pre-market systems, prototypes, academic projects, code, datasets, demonstrations, organizations, and development timelines.

The goal is not to repeat marketing claims. It is to show the entire field—from proven commercial products to early research—and make the differences clear.

A short history of household robots

The household robot has existed in popular imagination far longer than it has existed as a useful product.

Science-fiction stories and television programs established the idea of the robot housekeeper decades ago. The cultural image was usually humanoid: a machine that could move throughout the home, understand spoken instructions, manipulate ordinary objects, and perform many unrelated tasks.

The first successful household robots followed a very different path. Instead of attempting general intelligence, they performed one narrow task repeatedly.

The first robotic floor cleaners

Electrolux demonstrated its Trilobite robotic vacuum concept in the 1990s and commercially released it in 2001. It is generally recognized as the first commercially available autonomous robotic vacuum.

The more important mass-market moment came in 2002, when iRobot launched the Roomba. The early Roomba was far less sophisticated than today's models. It did not build the detailed visual and spatial maps now common in premium robots. Yet it solved an understandable household problem at a price consumers could accept.

That product established a pattern that still defines household robotics:

A robot succeeds in the home when it performs a frequent, undesirable task with enough reliability to justify its cost and maintenance.

The 2000s: navigation and acceptance

During the following decade, robotic vacuums became better at obstacle detection, charging, coverage, and floor transitions. Consumers also became accustomed to allowing autonomous machines to move around their homes.

This cultural acceptance mattered. A mobile robot with cameras, microphones, mapping software, and wireless connectivity would once have seemed highly unusual. Today, many households treat robotic cleaners as ordinary appliances.

The 2010s: mapping, apps, and specialization

The 2010s brought major improvements:

• LiDAR and vision-based mapping
• Smartphone control
• Room-level cleaning
• Scheduled operation
• Virtual boundaries
• Improved battery management
• Better obstacle avoidance
• Robotic lawn mowing
• More capable pool and window cleaning
• Social and educational robots

Household robotics began separating into recognizable categories rather than remaining synonymous with robotic vacuum cleaners.

The 2020s: self-maintenance, AI, and manipulation

Premium cleaning robots now commonly offer self-emptying dustbins, automatic mop washing, water refilling, mop drying, object recognition, improved threshold crossing, and increasingly sophisticated docking stations.

A significant transition is also beginning: movement from navigation-only robots to robots that can manipulate objects.

The Roborock Saros Z70, for example, adds a foldable mechanical arm to a mass-produced robotic vacuum. Its manipulation capability is limited to selected lightweight household items, but the concept is important. The robot is no longer merely navigating around an object; it can attempt to move it.

At the more ambitious end of the market, 1X has opened orders for NEO, a humanoid home robot designed to perform household tasks. Early units arrive with limited autonomy and can rely on scheduled remote expert supervision for tasks the robot cannot yet perform independently.

That distinction—between autonomous capability and human-assisted operation—is essential when evaluating the current market.

The household-robot market in numbers

No single organization maintains a complete global registry of every household robot, product variation, research prototype, crowdfunded system, and discontinued model.

Therefore, any product count must explain what it includes.

Official market evidence

The International Federation of Robotics reported approximately 20.1 million consumer service robots sold in 2024, an increase of 11 percent from the previous year in its reporting sample.

Most importantly, robots for domestic tasks represented approximately 97 percent of recorded consumer-service-robot sales.

By contrast, only 536 home-care robot sales were registered in the same statistical sample.

That contrast explains the current market better than almost any headline:

Household robots are already a high-volume market, but the overwhelming majority perform narrow domestic tasks. Complex care and general assistance remain niche.

The IFR also notes that its figures are based on a changing sample of suppliers rather than a projection of the entire industry. They should therefore be treated as strong market indicators, not as a complete global census.

How many types of household robots exist?

Using a practical taxonomy, the current field can be divided into approximately:

• 12–18 major commercial categories
• 30–40 meaningful subcategories
• Additional research-only and emerging categories

A useful top-level structure includes:

• Robotic floor cleaning
• Robotic lawn care
• Robotic pool cleaning
• Robotic window cleaning
• Other specialized cleaning
• Security and monitoring
• Telepresence
• Companion robots
• Robotic pets
• Educational robots
• Elder-care and assistance
• Mobility and feeding assistance
• Kitchen and food preparation
• Garden and outdoor maintenance
• General-purpose household assistants
• Humanoid household robots

How many household-robot products are actually sold?

Based on a broad market taxonomy and manufacturer catalog review, a reasonable editorial estimate is:

Approximately 700–1,300 distinct household-robot models are actively offered for purchase worldwide.

A broader count could reach:

Approximately 1,500–2,500 market listings

That broader number includes:

• Regional variants
• Rebranded or white-label products
• Products with minor model differences
• Crowdfunded products
• Limited-market releases
• Preorder systems
• Older models still appearing through retailers

A likely category distribution is:

These figures should not be interpreted as audited manufacturer totals. They are practical estimates designed to show the scale and shape of the market.

The most important fact is not whether the global total is 850 or 1,150. It is that a relatively small number of mature categories contain most commercially successful household robots.

The current household-robot landscape

1. Robotic vacuums and vacuum-mop systems

Market stage: fully commercial and mature

Robotic floor cleaning is the most successful household-robot category by a wide margin.

Modern premium systems can:

• Vacuum and mop
• Build detailed maps
• Identify rooms
• Recognize selected obstacles
• Avoid stairs
• Return to a charging dock
• Empty collected debris
• Refill clean water
• Wash and dry mop pads
• Adjust cleaning behavior by floor type
• Integrate with voice assistants and mobile apps

Leading examples

Roborock Saros series. The Saros 10 and Saros 10R represent the current premium direction: thin designs, sophisticated navigation, obstacle recognition, automated docks, and advanced mopping systems.

Roborock Saros Z70. The Z70 adds a five-axis mechanical arm capable of recognizing and moving selected lightweight objects. Its capabilities should not be confused with those of a general-purpose manipulator, but it represents an important bridge between cleaning robots and household manipulation.

iRobot Roomba Combo 10 Max. The Roomba Combo 10 Max combines vacuuming and mopping with a dock that empties debris and washes and dries the mop pad. Roomba remains historically important because it helped create the mass consumer-robot market.

Dreame premium vacuum-mop systems. Dreame's premium lines compete in mapping, obstacle handling, self-cleaning docks, suction, threshold navigation, and automated mop management.

Reality check

Even the best floor-cleaning robots still encounter: cables, small toys, pet waste, reflective or dark surfaces, complex furniture arrangements, tall thresholds, stairs, cluttered homes, and maintenance requirements.

They reduce household work; they do not eliminate it.

2. Robotic lawn mowers

Market stage: fully commercial and increasingly mature

Robotic lawn mowers have moved beyond simple boundary-wire systems. Newer models increasingly use satellite positioning, RTK navigation, computer vision, LiDAR, virtual boundaries, multi-zone mapping, all-wheel drive, obstacle recognition, and app-based route planning.

Leading examples

Husqvarna Automower 450X NERA. Designed for large lawns, the 450X NERA supports object avoidance and optional wire-free installation through satellite-based positioning.

Mammotion LUBA series. Mammotion's LUBA systems emphasize all-wheel drive, steep-slope capability, virtual boundaries, vision, and multi-zone operation. The lineup has expanded to address both large and smaller lawns.

Reality check

Robotic lawn mowers work best when the property has a manageable layout, edges and obstacles are mapped correctly, the lawn does not contain excessive debris, security and theft risks are controlled, and users understand seasonal storage and blade maintenance.

They are effective autonomous tools, not unattended landscape crews.

3. Robotic pool cleaners and surface skimmers

Market stage: fully commercial

Pool-cleaning robots are among the most practically valuable household robots because pool maintenance is repetitive, time-consuming, and well constrained.

Current products can clean combinations of pool floors, walls, waterlines, steps, and surface debris.

Leading examples

Maytronics Dolphin LIBERTY 600. This cordless cleaner supports floor, wall, waterline, and step cleaning, along with app control and scheduled eco-cleaning modes.

Beatbot AquaSense 2 Ultra. Beatbot markets the AquaSense 2 Ultra as a multi-function cordless pool cleaner using sensing and mapping to cover complex pool shapes.

Beatbot robotic skimmers. Surface-skimming robots provide a separate but complementary function, collecting leaves and floating debris before it sinks.

Reality check

Performance depends heavily on pool geometry, surface type, step design, water chemistry, filter maintenance, battery capacity, and debris type. "AI-powered" does not mean that every pool shape will be handled perfectly.

4. Window-cleaning robots

Market stage: commercial but less mature than floor cleaning

Window robots adhere to glass and move across the surface while wiping and spraying. They can be useful for large windows, high interior windows, exterior glass with safe access, and repetitive maintenance cleaning.

Leading examples

Ecovacs WINBOT W3 OMNI. The WINBOT W3 OMNI includes automated pad-cleaning support and a safety system designed to reduce fall risk.

HOBOT window-cleaning robots. HOBOT offers multiple established window-cleaning models using suction, navigation, and cleaning pads.

Reality check

Window-cleaning robots generally require more setup than robotic vacuums. Users may need to attach safety systems, place the robot manually, move it between panes, clean or replace pads, and monitor unusual glass arrangements. These are useful specialized machines, but not fully hands-off solutions.

5. Robotic companions and robotic pets

Market stage: commercially available but fragmented

Companion robots focus less on physical household labor and more on entertainment, emotional interaction, presence, communication, basic monitoring, personality, and education.

Leading examples

Sony aibo. Aibo remains one of the best-known robotic pets, combining movement, sensors, recognition, interaction, and an evolving personality experience.

Enabot EBO series. Enabot's mobile home robots combine communication, remote presence, monitoring, and companion-style behavior.

Loona. Loona is designed as an expressive robotic pet with visual recognition, movement, voice interaction, and personality-based behavior.

Reality check

The value of a companion robot is subjective. Buyers should examine subscription requirements, cloud dependence, long-term software support, camera and microphone privacy, repairability, and whether the interaction remains engaging after novelty fades.

6. Educational robots

Market stage: commercially established

Educational household robots help children and adults explore programming, logic, robotics, AI, creative problem solving, and science and engineering. Products range from screen-based social robots to programmable wheeled platforms and construction kits.

Representative products

Miko. Miko focuses on conversational and educational interaction for children.

Sphero and programmable robotics platforms. Sphero-style devices emphasize coding, movement, experimentation, and classroom or home learning.

LEGO robotics and related construction ecosystems. These systems remain valuable because they let users build, program, modify, and understand mechanisms rather than only interact with a closed finished product.

Reality check

Educational value depends more on curriculum, openness, and continued engagement than on how humanlike the robot appears.

7. Security, monitoring, and telepresence robots

Market stage: niche commercial

Mobile monitoring robots can let users check rooms remotely, speak with family members, observe pets, move a camera around the home, receive selected alerts, and provide basic telepresence.

Representative products

Enabot EBO X and related models. These products combine mobile monitoring, communication, navigation, and companion functions.

Temi. Temi has been used for telepresence, communication, and assistance, although its positioning spans both home and professional environments.

Reality check

A mobile camera inside the home introduces significant privacy and security questions. Buyers should assess how video is processed, whether data is stored in the cloud, who can access the device, whether software updates are guaranteed, and what happens if the manufacturer stops supporting the product.

8. Elder-care and assistive household robots

Market stage: early commercial, pilot, and research

This is one of the most socially important but technically difficult categories. Potential functions include medication reminders, social interaction, telepresence, fall or emergency detection, fetching objects, mobility support, feeding assistance, cognitive support, and routine monitoring.

Yet official sales data illustrates how limited this market remains compared with domestic cleaning.

Why is progress slower?

Care robots operate near vulnerable people and must meet much higher standards for safety, reliability, privacy, dignity, accessibility, human oversight, and error recovery.

A vacuum that misses a corner is inconvenient. A care robot that fails during a transfer or emergency can cause serious harm.

9. Kitchen and food-preparation robots

Market stage: specialized appliances are commercial; general robotic cooking remains early

The term "kitchen robot" covers very different products: automated cooking appliances, ingredient-processing machines, drink-making systems, robotic arms for food preparation, and research systems operating ordinary kitchens.

Some products automate cooking processes but do not possess the mobility and general manipulation associated with a household robot.

Reality check

Kitchens are especially difficult environments because they contain heat, water, knives, breakable objects, food contamination risks, highly variable containers, deformable materials, and cluttered workspaces. General robotic cooking remains far behind automated floor cleaning.

10. General-purpose and humanoid household robots

Market stage: preorder, pilot, prototype, and research

This category receives the most publicity and creates the most confusion.

A general-purpose household robot must combine reliable mobility, safe manipulation, object recognition, language understanding, long-horizon planning, memory, error recovery, human-aware behavior, navigation in clutter, and safe operation around children, pets, stairs, water, heat, and fragile objects.

These requirements are dramatically harder than vacuuming an open floor.

Important examples

1X NEO. NEO is one of the clearest efforts to introduce a humanoid robot specifically for household use. The system is orderable with a deposit, and 1X describes early access as providing foundational autonomy that should improve over time. For complex tasks, scheduled remote expert supervision may be used. This is not necessarily a weakness; human assistance can be a practical way to deploy emerging robots. But users should understand the difference between autonomous operation and remote human-supported operation.

Research mobile manipulators. Platforms such as Stretch, Fetch, PR2, TIAGo, and custom research robots are widely used to study household manipulation. A research system built on one of these platforms should not automatically be counted as a separate commercial robot product.

Reality check

The true household humanoid market is still in its earliest phase. A preorder does not equal large-scale deployment. A demonstration does not equal reliable everyday operation. A robot performing one chore in a controlled video does not mean it can perform that task across millions of different homes.

How Max Robotics will classify household robots

One reason household-robot coverage is confusing is that products, prototypes, methods, and research systems are frequently discussed as if they were equivalent.

Max Robotics will separate them by lifecycle stage.

Proposed lifecycle classification

1. Research concept — A paper, simulation, proposed system, or early experiment without a demonstrated complete physical robot.
2. Simulation-only system — A robot or task evaluated primarily in a virtual environment.
3. Laboratory prototype — A physical robot demonstrated under controlled laboratory conditions.
4. Functional prototype — A working system capable of completing meaningful tasks, but not yet deployed broadly.
5. Real-home test — A robot evaluated in one or more genuine residential environments.
6. Pilot or beta — A system deployed with selected users, households, partners, or test customers.
7. Crowdfunding — A product accepting financial backing but not yet established as a routinely delivered commercial product.
8. Preorder — A company is accepting reservations, deposits, or full orders for future delivery.
9. Limited release — A product is available only in selected markets, quantities, or customer programs.
10. Fully commercial — A product can be purchased and delivered through established sales channels.
11. Discontinued — A previously marketed product is no longer actively sold or supported.
12. Inactive or unknown — There is insufficient recent evidence to determine the project's status.

Each record should include original source links, status evidence, announcement or verification date, last checked date, related papers, organization, code, datasets, videos, household tasks, base platform, and confidence level.

This structure makes it possible to compare a mature robotic vacuum with an academic laundry-folding prototype without pretending that both occupy the same commercial stage.

How many household robots are still in research or pre-market development?

There is no authoritative total. A reasonable long-term collection estimate is:

• 100–300 active, visible company prototypes or announced household robots
• 200–600 pilot, beta, crowdfunding, limited-release, or pre-market systems
• 1,000–3,000 named academic household-robot systems
• 5,000–15,000 broader household-robot research projects, methods, configurations, and experimental platforms

These categories overlap and require deduplication.

For example, one household-manipulation system may produce multiple research papers, a dataset, a GitHub repository, a university project page, several demonstration videos, a startup, and a renamed commercial product. Those should not be treated as seven unrelated robots.

Max Robotics intends to connect them into one evidence-based lifecycle.

What household robots can really do today

Reliable or increasingly reliable: vacuuming, basic floor mopping, lawn mowing, pool-floor and wall cleaning, surface skimming, selected window cleaning, remote home monitoring, telepresence, entertainment and companionship, structured educational activities, repetitive reminders, and navigation within mapped environments.

Possible, but highly environment-dependent: recognizing household objects, avoiding clutter, moving selected lightweight objects, fetching known objects, opening selected doors or drawers, basic laundry handling, loading or unloading selected items, simple cleaning beyond floors, personalized conversation, and limited manipulation in prepared homes.

Still not generally solved: reliably cleaning an entire unmodified home, folding arbitrary laundry at human speed, washing and storing mixed dishes safely, cooking arbitrary meals in ordinary kitchens, handling transparent, deformable, sharp, hot, wet, and fragile objects, safely caring for people without meaningful human oversight, general household problem-solving across unfamiliar homes, operating for long periods without intervention, and recovering gracefully from every unexpected situation.

Why the home is one of robotics' hardest environments

Factories are designed around automation. Homes are designed around people.

Homes contain unpredictable layouts, moving people, pets, children, narrow passages, stairs, soft objects, reflective objects, transparent objects, clutter, poor lighting, constantly changing item locations, personal belongings, and privacy-sensitive areas.

A factory robot can repeat one motion in a controlled cell. A household robot must understand a constantly changing world and operate safely around untrained users.

This is why specialized household robots have reached mass adoption while general-purpose systems have not.

How AI is changing the course of household robots

For most of their history, household robots were not "intelligent." They were carefully engineered machines running hand-written rules. A robot vacuum followed mapping and coverage algorithms. A lawn mower followed a boundary. Almost none of them understood language, reasoned about unfamiliar situations, or generalized to tasks they were not explicitly programmed for.

The recent wave of large language models and foundation models is beginning to change that—though not yet in the way the most dramatic headlines suggest.

What the new AI actually contributes

Models such as ChatGPT (OpenAI), Claude (Anthropic), Gemini (Google), and DeepSeek were built to process language, not to fold laundry. But the capabilities they demonstrate—understanding instructions, breaking a goal into steps, reasoning about everyday situations, and explaining what they are doing—are exactly the capabilities household robots have always lacked.

Three shifts matter most:

• Language understanding. A person can say "tidy the living room before guests arrive" instead of selecting a preset. The model interprets intent.
• Planning and common sense. Foundation models can decompose a vague goal into ordered steps and bring background knowledge about how homes and objects usually work.
• Generalization. Instead of programming every object and situation in advance, a model trained broadly can attempt tasks and objects it has never explicitly seen.

A second, robotics-specific family of models is also emerging: vision-language-action (VLA) models that connect perception and language directly to physical motion—Google's RT-2 and Open X-Embodiment / RT-X, NVIDIA's Project GR00T for humanoids, and startups such as Physical Intelligence (π0) and Skild AI, which recently raised one of the largest robotics rounds on record. These are the systems trying to turn "understanding a task" into "performing it with a real body."

What already uses this today

The technology is no longer purely experimental:

• Humanoid developers are the most aggressive adopters. Figure has worked with OpenAI's models for language and visual reasoning and has since built in-house vision-language-action models; 1X NEO runs learned neural policies rather than scripted routines; and Boston Dynamics has said its new electric Atlas program involves Google DeepMind as a partner.
• Companion and educational robots already embed conversational AI. Products like Miko use large-language-model-style dialogue, and consumer concepts such as Samsung's Ballie have been demonstrated with generative AI for natural conversation and home control.
• Mainstream appliances are starting to add it at the edges. Some premium robot vacuums now ship large-language-model voice assistants for natural commands and questions, layered on top of their existing navigation.
• Chinese manufacturers increasingly build on open-weight models such as DeepSeek and Qwen, which lower cost and allow more on-device or self-hosted intelligence—an important factor in a price-sensitive, high-volume market.

It is worth being precise: in most of these products, the AI improves the interface and reasoning layer—understanding what you want and planning—while the physical work still relies on conventional, task-specific engineering.

How it is likely to evolve

The most probable trajectory is not "ChatGPT suddenly does the dishes." It is a gradual fusion of language-level intelligence with physical capability:

1. Natural-language control of existing specialized robots.
2. Foundation models that plan multi-step household chores.
3. VLA models that map instructions directly to manipulation—first for standardized objects, then for messier ones.
4. On-device models that run locally for privacy, latency, and reliability.
5. Robots that learn from demonstration and from each other's data, improving after they ship.

The honest limitation

Language was never the hard part of housework.

A model that can write an essay about cleaning a kitchen is not the same as a robot that can clean one. The unsolved problems—reliable manipulation of unfamiliar objects, safe operation around people, recovery from mistakes, and consistent performance across millions of different homes—are physical and safety problems, not language problems.

Large language models make robots far easier to talk to and far better at planning. They do not, on their own, make a robot's hands more capable. The "ChatGPT moment" for household robots will arrive when language-level intelligence is reliably grounded in physical action—and the evidence suggests that moment is being approached in stages, not in a single leap.

The next five to ten years

1. Specialized robots will continue to dominate. Robotic vacuums, lawn mowers, pool cleaners, and other task-specific products will keep improving because their economics and use cases are already proven. Expect better obstacle handling, more self-maintenance, improved mapping, lower prices, better energy management, more capable docks, gradual manipulation features, and stronger interoperability.

2. Manipulation will expand from small controlled tasks. The progression will probably not jump directly from vacuum cleaners to universal humanoid housekeepers. A more likely sequence is: navigate around objects → identify objects → move selected lightweight objects → pick up standardized items → open selected doors and drawers → perform structured multi-step chores → generalize across more homes and objects. Robotic vacuums with small arms may look like a novelty, but they are also an early test of this progression.

3. Remote assistance will remain important. Some early general-purpose household robots will combine autonomous operation with remote human support. This creates practical value before complete autonomy is achieved, but it also raises questions: Who can see into the home? When is remote access activated? How is consent handled? Where is video stored? Can the owner verify whether a human is controlling the robot? What happens if network access fails? Transparent labeling will matter.

4. The market may favor multiple forms, not only humanoids. Humanoid robots receive attention because homes are designed for human bodies. However, the winning household robot may not always resemble a person. Different tasks may favor low mobile bases, mobile manipulators, wall-mounted arms, ceiling systems, small companion robots, smart appliances, wearable assistance, or cooperative groups of devices. The future home may contain several specialized robots before it contains one universal humanoid.

5. Research-to-product timelines will become valuable. The most important intelligence may come from observing movement across stages: paper → code → dataset → prototype → home trial → startup → preorder → commercial product. That chain helps answer which research areas are becoming commercially viable, which companies are translating academic work into products, which household tasks have many papers but few products, which prototypes are closest to market, which commercial claims have strong research support, and which systems have been tested in real homes.

Why Max Robotics is building this resource

The household-robot field does not need another page filled with promotional videos and futuristic claims. It needs an evidence-based resource that connects research papers, methods, authors, institutions, technologies, datasets, code repositories, products, companies, videos, lifecycle stages, commercial availability, historical timelines, and real-world limitations.

Max Robotics already supports natural-language discovery across approximately 500,000 robotics and AI papers. Users can search by meaning rather than only exact wording. A question about "fast household object manipulation using inexpensive cameras" can surface research that uses related language such as real-time perception, RGB-D manipulation, low-latency inference, or affordable vision systems.

The Research Advisor—already live and open to the public—goes beyond returning paper titles. It helps users investigate which methods are most relevant, which approaches have code, which datasets are commonly used, which systems were tested on real robots, which methods are fastest, most practical, or easiest to implement, and what limitations appear repeatedly across the literature.

The planned Product Advisor will bring the same plain-English experience to household-robot products and development-stage systems. Users should eventually be able to ask:

• What are the best commercially available robots for cleaning large windows?
• Which robotic lawn mowers work without boundary wires?
• Which household robots have actually been tested in real homes?
• Which humanoid home robots can be ordered today?
• Which products are fully autonomous and which depend on remote assistance?
• Which research prototypes are closest to commercialization?
• Which household tasks have strong research but no mature products?
• What is the difference between a commercial robot, a preorder, and a laboratory demonstration?

The site will not need to host copyrighted papers, videos, or datasets. It can provide structured information, abstracts, independent summaries, status labels, and links to original sources.

The bottom line

Household robotics is simultaneously mature and immature.

It is mature where the task is narrow, the environment is constrained, the value is easy to understand, failures are inconvenient rather than dangerous, and the robot can repeat the same job.

It is immature where tasks require general manipulation, environments change constantly, safety is critical, human behavior must be understood, privacy is sensitive, and errors have serious consequences.

The market already sells millions of household robots each year, but those robots are overwhelmingly specialized. The universal robotic housekeeper remains an emerging system rather than an established appliance.

That does not make the future less exciting. It makes accurate classification more important.

Max Robotics aims to become the place where people can see the entire picture: what is commercially proven, what is newly released, what is accepting preorders, what is being piloted, what remains a laboratory prototype, and which ideas exist only in research.

In a field filled with hype, the most useful resource will not be the one that makes the biggest promise. It will be the one that shows the strongest evidence.

Sources & fact notes

The IFR's 2025 executive summary reports 20.1 million consumer service robots sold in 2024, 11% annual growth, and says domestic-task robots represented 97% of consumer-service-robot sales. It recorded only 536 home-care robot sales in the same sample. IFR also cautions that its figures are based on supplier samples and should not be treated as projections of the entire industry.

IFR defines consumer robots as service robots intended for laypeople who do not require professional training, including domestic cleaners, automated wheelchairs, and social-interaction robots.

The historical discussion is based on the Electrolux Trilobite's 1990s demonstration and 2001 commercial release, followed by the Roomba's 2002 launch and mass-market success.

Current floor-cleaning examples and capabilities are drawn from official product information for the Roomba Combo 10 Max, Roborock Saros 10R, Saros 10, and Saros Z70. Roborock describes the Z70 as a mass-produced robotic vacuum with a five-axis arm and limits its current organizing behavior to selected lightweight objects.

The lawn-mower examples are based on official Husqvarna and Mammotion materials. Husqvarna specifies lawns up to 5,000 square meters for the 450X NERA, while Mammotion describes virtual mapping, all-wheel drive, vision, LiDAR, and obstacle detection across current models.

The pool-cleaner examples and functions are based on official Maytronics and Beatbot product information.

The current window-cleaning example is based on Ecovacs' official WINBOT W3 OMNI information.

The description of 1X NEO is based primarily on 1X's official product page. It states that early owners receive foundational autonomy, that capabilities are expected to grow, and that scheduled Expert Mode may remotely supervise complex tasks. The current order page lists a $200 deposit.

The estimates for active products, pre-market systems, and research projects are Max Robotics editorial planning estimates, not official industry totals. They should be updated as the site's product and research-system registry grows.