Humanoid robots are moving from science-fiction demos into real industrial planning. Xiaomi CEO Lei Jun’s recent prediction that humanoid robots will soon leave the lab and begin working at scale in factories captures a major shift in automation: robots are no longer being designed only for isolated cages, fixed assembly lines or single-purpose machines. The new wave of humanoid robots is being built to operate in spaces originally designed for humans — factory floors, hospital corridors, logistics centres, retail backrooms and eventually homes.
This does not mean that humanoid robots will replace all workers overnight. The more realistic future is uneven: robots will first take over repetitive, physically demanding, dangerous and highly structured tasks. Human workers will continue to dominate roles requiring judgment, empathy, supervision, negotiation, creativity and responsibility. But the direction is clear. As artificial intelligence improves robot perception, motion planning and language understanding, the labour market will face one of the biggest changes since the rise of industrial automation.
Why Humanoid Robots Are Becoming Practical Now
Humanoid robots were difficult to commercialize for decades because they required a combination of advanced motors, batteries, sensors, balance control, real-time computing and reliable software. What has changed is the convergence of several technologies.
First, AI models have improved dramatically. Vision-language models can help robots understand objects, instructions and environments. Second, hardware costs are falling as China, the United States, Japan and Europe invest heavily in robotics supply chains. Third, labour shortages in manufacturing, logistics, elder care and healthcare are pushing companies to search for alternatives. Fourth, industries already using automation now want more flexible machines that can work with existing tools and infrastructure instead of requiring complete factory redesign.
Humanoid robots are attractive because they are shaped for a human-built world. A robot with arms, hands, legs and a human-like height can potentially use doors, shelves, tools, stairs, carts and workstations already designed for people. This makes the humanoid form especially interesting for companies that do not want to rebuild their entire workplace around robots.
Sectors Adopting Humanoid and Advanced Robots
1. Manufacturing and Automotive
Manufacturing is the first major adoption zone. Lei Jun’s prediction focuses on Xiaomi factories, especially as the company expands from smartphones and electronics into electric vehicles and smart manufacturing. Automotive factories are particularly suitable because they already use automation, have measurable workflows and face pressure to improve efficiency.
Humanoid robots are being tested for tasks such as moving components, loading parts, inspecting products, carrying tools, sorting materials and assisting assembly-line workers. These jobs are often repetitive and physically demanding. In the short term, robots are more likely to assist workers than fully replace them. In the medium term, some roles involving simple material handling and repetitive station work may shrink.
The biggest opportunity is not just “robots replacing workers,” but robots making factories more flexible. Traditional industrial robots are excellent at fixed repetitive tasks, but humanoids could be redeployed from one task to another with software changes, making them useful in factories with changing product designs.
2. Logistics and Warehousing
Logistics is another leading sector. Warehouses and fulfilment centres are full of repetitive physical tasks: moving totes, scanning items, unloading carts, sorting parcels and transporting goods between stations. Companies such as Amazon and Agility Robotics have already tested humanoid or human-centric robots for warehouse work.
The reason logistics is attractive is simple: warehouses are semi-structured environments. They are more predictable than streets or homes, but more flexible than traditional assembly lines. Robots can learn repeated routes, handle standard bins and work alongside employees.
For employment, logistics may see a gradual reduction in low-skill repetitive roles, especially night-shift and heavy-lifting jobs. But new jobs will appear in robot supervision, fleet maintenance, warehouse robotics operations, safety compliance, data labelling, simulation training and workflow redesign.
3. Healthcare and Hospitals
Hospitals are a high-potential but high-risk sector. Robots can help with non-clinical work such as moving supplies, delivering medicines, transporting linens, cleaning rooms and supporting elderly patients. In the future, humanoid robots may assist nurses, support rehabilitation, help with patient monitoring and perform teleoperated medical tasks.
However, hospitals are more complicated than factories. Safety, regulation, hygiene, liability and patient trust are major barriers. A robot working in a warehouse can make a mistake with a box; a robot working near a patient must meet much higher standards.
The near-term opportunity in hospitals is not replacing doctors or nurses. It is reducing the burden of repetitive support tasks so healthcare workers can spend more time on patient care. This is especially important in countries facing ageing populations and healthcare labour shortages.
4. Elder Care and Domestic Assistance
Elder care may become one of the largest long-term markets. Many countries face ageing populations and a shortage of caregivers. Humanoid robots could help elderly people with reminders, mobility support, companionship, simple household chores and emergency alerts.
But this sector will take longer to mature. Homes are unpredictable. A factory floor is organized; a home has pets, children, clutter, carpets, stairs, different lighting and emotional complexity. Robots will need better dexterity, safety and social intelligence before they become common in homes.
5. Retail, Hospitality and Public Services
Retail and hospitality robots may serve as guides, cleaners, stock assistants, hotel delivery workers or reception helpers. These environments are public-facing, so design, communication and trust matter. Robots must not only perform tasks but also interact naturally with people.
Adoption here will depend heavily on customer acceptance. In some cultures, people may welcome robots; in others, customers may prefer human service. The best use cases will be behind-the-scenes tasks such as inventory checking, shelf scanning, cleaning and internal delivery.
Implications for Employment
The biggest labour-market impact will be task displacement, not immediate job extinction. Most jobs are bundles of tasks. Robots may take over some tasks while humans continue doing others. For example, in a hospital, a robot may deliver supplies, but a nurse will still handle judgment, empathy and patient communication. In a factory, robots may move parts, but technicians will supervise quality, repair machines and solve process problems.
The workers most exposed are those in repetitive, physical and predictable roles. These include material handlers, warehouse associates, basic assembly workers, cleaners, porters and some logistics staff. Jobs requiring human judgment, emotional intelligence, technical troubleshooting and complex decision-making are less vulnerable.
There will also be new employment categories. Companies will need robot maintenance technicians, robotics safety managers, fleet operators, AI trainers, simulation engineers, data specialists, robot integration consultants and human-robot workflow designers. The challenge is that displaced workers may not automatically qualify for these new roles without training.
Governments and companies must prepare early. Reskilling programs should focus on practical technical skills: robotics maintenance, electrical basics, machine vision, warehouse automation systems, safety procedures, AI operations and human-machine collaboration. The countries that manage this transition well will gain productivity without creating extreme social disruption.
Risks and Challenges
Humanoid robots still face serious technical and economic challenges. Dexterous hands remain difficult. Battery life is limited. Robots must become safer around humans. Maintenance costs can be high. Many impressive demonstrations are still controlled or teleoperated rather than fully autonomous.
There is also a risk of overhype. Some companies may market humanoid robots before they are ready for real-world deployment. Investors should separate demonstration videos from actual commercial performance. The key question is not whether a robot can walk or dance, but whether it can perform useful work reliably for thousands of hours at a cost lower than human labour or traditional automation.
Cybersecurity is another concern. Humanoid robots may carry cameras, microphones, sensors and network connections into sensitive workplaces. In factories, hospitals and logistics centres, robot data security will become a major policy issue.
Investment Opportunities in Robotics
The robotics opportunity is broader than buying shares in humanoid robot makers. The value chain includes hardware, software, components, AI infrastructure, sensors, batteries, semiconductors, simulation platforms and robot-as-a-service business models.
Hardware Opportunities
Key hardware areas include actuators, motors, gearboxes, robotic hands, force sensors, batteries, lightweight materials and machine-vision cameras. Companies that supply critical components may benefit even if no single humanoid robot brand dominates the market. Robotic hands are especially important because manipulation is one of the hardest problems in robotics.
Software Opportunities
Software may become the highest-margin layer. Robots need operating systems, fleet management, simulation environments, safety monitoring, visual perception, reinforcement learning, navigation and task-planning software. As robots become more common, companies will pay for platforms that train, monitor and update robot fleets.
AI and Chips
Humanoid robots require edge computing, GPUs, vision processors and real-time control chips. AI hardware companies may benefit from robotics demand, especially if robots require local inference rather than constant cloud connection.
Industrial Integration
Many factories cannot simply buy a robot and switch it on. They need integration, workflow redesign, safety audits and maintenance contracts. This creates opportunities for robotics integrators, automation consultants and industrial software providers.
ETFs and Diversified Exposure
For ordinary investors, robotics and AI exchange-traded funds may offer diversified exposure to automation, industrial robotics, AI chips and related software companies. This reduces the risk of betting on one humanoid startup that may fail. However, robotics ETFs can still be volatile because many holdings are tied to high-growth technology cycles.
The Most Likely Future
Humanoid robots will not suddenly replace all human labour, but they will steadily change the structure of work. The first major impact will be in factories and logistics centres, where tasks are repetitive, environments are controlled and return on investment is easier to measure. Hospitals and elder care will follow more slowly because safety and trust requirements are higher.
Lei Jun’s prediction is important because it reflects a broader industry belief: the humanoid robot is becoming a practical labour platform. The next decade will decide whether humanoids become a niche tool or a general-purpose workforce technology.
The winners will be companies that solve real problems, not just produce viral robot videos. The biggest opportunities will come from reliable hardware, scalable software, safe deployment and measurable productivity gains. For workers, the message is equally clear: the future will not be human versus robot, but humans who can work with robots versus those who are left unprepared for the automation shift.
Humanoid robots are leaving the lab. The real question is not whether they will enter the workplace, but how quickly businesses, workers and governments can adapt.