Humanoid robots are no longer limited to research labs and technology shows.

Some are now lifting parts, moving containers, handling materials, and working inside real factories and warehouses.

BMW, Hyundai, Mercedes-Benz, GXO Logistics, and other major companies are testing or deploying humanoid robots for practical work.

This raises a much bigger question.

Will humanoid robots eventually become normal coworkers?

The short answer is that they are already starting to enter workplaces. However, widespread adoption will depend on safety, reliability, cost, artificial intelligence, and whether robots can provide real value for businesses.

Here is what is happening now and what it could mean for the future of work.

What Is a Humanoid Robot?

A humanoid robot is a machine designed with some features similar to the human body.

Most have a torso, arms, hands, legs, cameras, sensors, and computer systems that help them understand their surroundings.

Some walk on two legs. Others use wheels for easier movement.

Their human-like shape has an important purpose.

Most workplaces were designed for people.

Doors, stairs, shelves, workstations, tools, boxes, and production lines are usually built around human height and movement.

A humanoid robot may therefore be able to work in an existing building without forcing a company to redesign the whole workplace.

That could be one of the technology’s biggest advantages.

Humanoid Robots Are Already Doing Real Work

The humanoid robotics industry has spent years producing impressive demonstrations.

The situation is now changing.

Several robots have moved from controlled demonstrations into factories and logistics facilities.

Here are some notable examples.

Humanoid Robot	Company	Workplace	Main Tasks	Current Development
Atlas	Boston Dynamics	Hyundai manufacturing operations	Part sequencing, material handling and future assembly work	Training and industrial integration
Figure 03	Figure AI	BMW Plant Spartanburg	Logistics, material movement and sequencing	Factory deployment and testing
Digit	Agility Robotics	GXO and other industrial sites	Moving totes and warehouse materials	Commercial industrial deployment
Apollo	Apptronik	Mercedes-Benz production environments	Intralogistics, material movement and inspection	Testing and training
AEON	Hexagon	BMW Plant Leipzig	Battery production and manufacturing tasks	Industrial pilot

The important change is not how human these robots look.

It is whether they can perform useful work for hours at a time.

BMW Has Already Put Humanoid Robots to Work

BMW provides one of the clearest examples of humanoid robots moving beyond demonstrations.

Figure AI deployed its Figure 02 robot at BMW’s Spartanburg plant in the United States.

According to Figure, the robot operated during 10-hour shifts, five days a week. It loaded more than 90,000 parts and recorded more than 1,250 operating hours.

Figure says the deployment contributed to the production of more than 30,000 BMW X3 vehicles. Read Figure AI’s BMW deployment report

Figure 02 has since been retired as Figure moves to its newer generation.

In June 2026, Figure 03 arrived at BMW’s Spartanburg facility. The newer robot is being used to explore more complex logistics work, including sequencing parts and moving equipment. See Figure 03 working at BMW

BMW is also expanding its humanoid robot experiments beyond the United States.

In September 2026, the company announced a humanoid robotics project at its Leipzig plant in Germany. The project uses the AEON robot for work related to battery module production and other manufacturing processes. Read BMW’s humanoid robotics update

These projects show how quickly industrial humanoid robotics is developing.

Boston Dynamics Is Preparing Atlas for Factory Work

Boston Dynamics is famous for videos showing robots running, jumping, and performing complex movements.

Its latest Atlas robot has a more practical goal.

It is being developed as an industrial humanoid.

In September 2026, Boston Dynamics opened the Robotics Metaplant Application Center at Hyundai Motor Group’s manufacturing campus in Georgia.

Atlas robots are being trained there using real manufacturing environments.

One early task involves preparing and sequencing automotive parts before they reach the assembly process. Explore Boston Dynamics’ Atlas factory program

Hyundai has outlined a gradual plan for Atlas.

The company has said it expects Atlas to begin entering validated manufacturing processes such as parts sequencing from 2028. It expects the range of tasks to expand later.

This gradual approach matters.

Companies cannot simply place an advanced robot inside a factory and expect it to perform every human job.

Each task must be tested for safety, speed, accuracy, and cost.

Digit Is Moving Thousands of Items in Warehouses

Agility Robotics has taken a different approach with its Digit humanoid robot.

Digit was designed around practical industrial work.

Its commercial deployment at a GXO facility near Atlanta has focused on material handling.

The robot receives containers and moves them into the next stage of the warehouse process.

Agility reported in 2026 that Digit had moved more than 100,000 totes at the GXO facility.

The company also says earlier versions of Digit accumulated more than 65,000 hours of operation across commercial deployments and industrial sites.

Its newer Digit 5 model has been designed to work more closely around people without the large physical barriers required by many traditional industrial robots. See Agility Robotics’ Digit 5 announcement

This type of repetitive material handling may become one of the first major markets for humanoid robots.

Mercedes-Benz Is Testing Apollo

Mercedes-Benz is also exploring humanoid robots.

The company has been working with Apptronik and its Apollo robot.

Apollo is being tested for production and intralogistics tasks.

That can include moving components to production lines and carrying out basic inspections.

Mercedes-Benz employees have also helped train the system through teleoperation and augmented reality.

The goal is to teach Apollo how manufacturing tasks are performed before allowing the robot to complete more work autonomously. Read about Mercedes-Benz and Apollo

This points to an important part of the humanoid robot story.

Humans may not simply work beside robots.

They may also teach them.

Why Are Companies Building Human-Shaped Robots?

Factories already contain huge numbers of robots.

So why do companies need humanoid ones?

Traditional industrial robots are excellent when performing a fixed task.

A robotic arm may weld the same part thousands of times with extreme accuracy.

The problem is flexibility.

If a company wants that robot to perform a completely different job, it may need new hardware, new safety barriers, new software, and changes to the production line.

Humanoid robots are designed around a different idea.

One machine could eventually learn many tasks.

The same robot might move boxes in the morning, load parts in the afternoon, and perform inspections later.

That level of flexibility has not yet been achieved at large scale.

However, it is one of the main reasons companies are investing heavily in humanoid technology.

Physical AI Is Making Robots Smarter

Hardware alone is not driving this change.

Artificial intelligence is just as important.

The term physical AI describes AI systems that can understand and interact with the physical world.

A chatbot mainly works with digital information.

A humanoid robot must deal with reality.

It needs to recognize objects.

It must understand distance.

It needs to know how much force to use.

It must balance while walking.

It may need to decide how to pick up an object it has never seen before.

It also has to react when something unexpected happens.

Modern AI models, computer vision, reinforcement learning, simulation, and large amounts of robot training data are helping improve these abilities.

This is why the progress of AI and humanoid robotics is becoming closely connected.

What Jobs Will Humanoid Robots Do First?

Humanoid robots are unlikely to begin with complicated jobs that require constant judgment.

The first large deployments are more likely to focus on structured physical tasks.

Material handling is a strong example.

A robot can move containers, carry parts, load equipment, transport components, or place objects in known locations.

Repetitive manufacturing tasks are another target.

Companies are also interested in jobs that involve heavy lifting or uncomfortable movements.

These tasks can cause physical strain when people repeat them for hours.

Dangerous environments could become another important use case.

Robots may eventually enter areas involving heat, chemicals, unstable structures, or other risks before sending human workers inside.

What Are the Benefits of Humanoid Robots?

The biggest potential benefit is flexibility.

A general-purpose robot could complete different tasks without requiring a completely new machine for each job.

Another benefit is workplace safety.

Robots can perform repetitive lifting and other physically demanding activities.

They can also enter environments that may be unsafe for people.

Availability is another factor.

Machines do not need normal meal breaks or sleep.

However, robots still need charging, maintenance, repairs, software updates, and inspections.

Productivity could also improve when robots and humans divide tasks based on their strengths.

A robot might perform repetitive material movement while a human handles quality control, problem solving, communication, or complex decisions.

The Biggest Problems Humanoid Robots Still Need to Solve

Humanoid robots look impressive, but real workplaces are difficult.

Reliability

A factory needs machines that work consistently.

A robot that performs perfectly in a five-minute demonstration may still struggle during a ten-hour production shift.

Businesses care about uptime more than viral videos.

Safety

A humanoid robot is large, heavy, and powerful.

That creates serious safety requirements.

Robots need sensors, emergency systems, movement limits, collision detection, and clear operating procedures.

This becomes even more important when robots and people share the same area.

Battery Life

Walking, balancing, lifting, computing, and using sensors consume energy.

Companies therefore need robots that can work long enough to provide economic value.

Fast charging and replaceable batteries may help.

Dexterity

Human hands are extremely difficult to copy.

People can pick up a bolt, open a package, connect a cable, use a screwdriver, and handle fragile objects without thinking much about it.

For robots, these tasks can be difficult.

Better robotic hands and AI control systems are therefore critical.

Cost

A humanoid robot only makes business sense if the value it produces is greater than the total cost of owning and operating it.

That calculation includes the robot itself, maintenance, electricity, software, integration, support, training, and downtime.

Lower hardware costs will make wider adoption easier.

Will Humanoid Robots Replace Human Jobs?

This is one of the biggest concerns surrounding workplace robotics.

The likely impact is more complex than simply replacing entire occupations.

Robots normally automate tasks rather than complete jobs.

A warehouse employee may spend part of the day moving containers and another part checking inventory, solving problems, communicating with coworkers, and handling unexpected situations.

A robot may automate the container movement without performing every other part of that person’s role.

Research highlighted by the International Federation of Robotics describes several effects of workplace automation.

Robots can substitute for some routine tasks. They can also create complementary work involving installation, programming, maintenance, supervision, engineering, and process design. Explore research from the International Federation of Robotics

The actual effect will differ between industries, companies, countries, and occupations.

New Jobs Could Appear Around Robot Workforces

If companies deploy hundreds or thousands of humanoid robots, those machines will need support.

Someone will need to maintain them.

Someone will monitor their performance.

Technicians will repair damaged hardware.

Engineers will improve workflows.

Safety specialists will study how humans and robots interact.

AI teams will train robots for new tasks.

Companies may also employ robot fleet managers who supervise large groups of machines.

Humanoid robots could therefore remove some tasks while creating demand for different skills.

Worker training will be an important part of that transition.

Could Humanoid Robots Enter Offices and Retail Stores?

They could, but factories and warehouses currently offer a clearer business case.

Industrial environments contain large amounts of repetitive physical work.

Tasks can also be measured easily.

A company can calculate how many containers a robot moves per hour or how many components it loads during a shift.

Retail is less predictable.

A robot inside a store may need to move through crowds, understand customers, handle many different objects, and respond safely to unexpected behavior.

Office environments have another problem.

Most office work happens on computers.

Software-based AI agents may be cheaper and more practical than physical humanoid robots for many office tasks.

For these reasons, factories and logistics facilities are likely to remain important testing grounds for humanoid technology.

Humanoid Robots vs Traditional Industrial Robots

Feature	Humanoid Robot	Traditional Industrial Robot
Flexibility	Designed for multiple tasks	Usually optimized for specific tasks
Mobility	Can move between areas	Often fixed in one location
Human environments	Designed to use human spaces	May require dedicated infrastructure
Task changes	Potentially software-driven	Often requires hardware or workflow changes
Maturity	Still developing rapidly	Highly mature
Reliability	Improving	Usually very high
Cost efficiency	Still being proven	Well understood for established tasks

Traditional robots will not suddenly disappear.

They remain faster and more efficient for many repetitive industrial tasks.

Humanoid robots make more sense when flexibility and mobility provide a clear advantage.

When Will Humanoid Robots Become Normal Coworkers?

There is no single date.

The change is likely to happen industry by industry and task by task.

Manufacturing and logistics are already providing early evidence.

The next stage will depend on whether robots can operate reliably for long periods while delivering a clear financial benefit.

Safety must also improve.

Robots need to work around people without creating new workplace risks.

Cost will matter too.

A technically impressive robot will not become common if using it costs far more than existing automation.

The transition may therefore feel gradual.

A company may start with one robot.

Then it may deploy five.

Those robots could initially perform one task.

Later, software updates may allow the same machines to handle several tasks.

Eventually, seeing humanoid robots inside large factories and warehouses may stop feeling unusual.

Final Thoughts

Humanoid robots are crossing an important line.

For years, the industry was dominated by impressive research projects and demonstrations.

Now companies are asking a harder question.

Can these machines perform useful work every day?

BMW’s experience with Figure, GXO’s use of Digit, Mercedes-Benz’s work with Apollo, and Hyundai’s preparations for Atlas suggest that companies are serious about finding the answer.

There are still major obstacles.

Humanoid robots need better reliability, safer operation, longer working times, stronger dexterity, and a convincing return on investment.

They also need to prove that their human-like design provides advantages over traditional automation.

For now, humanoid robots are unlikely to appear beside every worker.

Factories and warehouses are a different story.

Those environments are quickly becoming the testing ground for a new type of workforce where people, artificial intelligence, traditional automation, and humanoid machines work together.

The most important question may no longer be whether humanoid robots can walk like humans.

It is whether they can become useful enough, safe enough, and affordable enough to earn a permanent place at work.

Frequently Asked Questions

Are humanoid robots already working in factories?

Yes. Humanoid robots have already been tested and deployed in real manufacturing and logistics environments. Examples include Figure robots at BMW, Digit at GXO, Apollo with Mercedes-Benz, and Atlas training within Hyundai’s manufacturing ecosystem.

What are humanoid robots used for at work?

Current use cases include moving materials, handling containers, loading parts, sequencing components, assisting production processes, and performing repetitive physical tasks.

Why make robots look like humans?

Most workplaces are designed around human movement and proportions. A human-shaped robot may be able to use existing doors, shelves, tools, workstations, and production areas without major changes to the building.

Will humanoid robots replace workers?

They may automate certain tasks, especially repetitive or physically demanding ones. Their wider employment impact will depend on the industry and how companies redesign jobs around automation.

Are humanoid robots powered by AI?

Most modern humanoid systems combine robotics with artificial intelligence. AI can help with vision, navigation, object recognition, movement planning, manipulation, and learning new tasks.

Are humanoid robots safe around people?

Safety is still one of the industry’s most important challenges. Companies are developing collision detection, sensors, movement controls, emergency systems, and new operating procedures to allow robots to work more safely near humans.

Which companies are developing humanoid robots?

Major developers include Boston Dynamics, Figure AI, Agility Robotics, Apptronik, Tesla, Unitree, 1X, Sanctuary AI, and several other robotics companies.

What industries could use humanoid robots?

Manufacturing and logistics are currently major areas of development. Future uses could include construction, healthcare support, retail, hospitality, maintenance, agriculture, and work in hazardous environments.