%page_title%

Imagine a scene straight out of a science fiction film, with machines working alongside operators in production halls. These machines are not only precise and continuous in their work, but they can also look people in the eye, react to emotions, and respond in ways that, until recently, seemed to be the preserve of literature and cinema. This is not a futuristic vision, but an increasingly tangible reality: the reality of humanoid robots.

We are living in an era in which the distinction between humans and machines is becoming increasingly blurred, with engineering and artificial intelligence setting new directions for the development of industry, medicine, and everyday social interactions. Equipped with AI algorithms, sensors and actuators, humanoid robots are entering areas where monotony and risk are driving humans out, but where empathy and human-like communication are also needed. They demonstrate that digital intelligence takes on a new dimension when given physical form with the capacity for cooperation and empathy.

This story did not begin yesterday. As early as 1972, WABOT-1, the first full-size humanoid capable of walking and grasping, was created at Waseda University in Japan. However, the real breakthrough came in 2000 with Honda’s ASIMO — a robot that could walk, climb stairs and interact with people. Since then, the pace of innovation has grown at a dizzying rate. Today, Atlas from Boston Dynamics amazes us with its acrobatic skills, and Ameca, which was created in England, is becoming a symbol of how machines can form social and entertainment-based relationships.

What was once considered a special effect is now entering hospitals, factories, and public spaces. These are no longer experiments confined to laboratories, but rather the first steps of a technological revolution that we are witnessing. Not only may humanoid robots become the answer to staff shortages in many sectors, they may also be a catalyst for deeper questions about the future of work, the role of humans, and the boundaries between the biological and artificial worlds.

Although engineers and scientists shape them, their presence is increasingly influencing the shape of society as a whole. The question we should be asking today is whether we are teaching machines to be more human or if they will force us to redefine what it means to be human.

Humanoid robots: market data and analysis

When we look back at industrial history, we see that every technological revolution began as a response to a pressing problem of the time. The steam revolution freed economies from the constraints of muscle power; electrification introduced a new pace of production; and 20^(th)-century automation delivered unprecedented efficiency. Today, however, the global economy is facing a challenge that no longer fits within the boundaries of a single industry or continent: a shortage of labour. Demographic forecasts paint a bleak picture: by 2050, Europe’s working-age population is expected to shrink by 10%. Meanwhile, inflationary pressure, productivity stagnation, and the forced regionalisation of supply chains are prompting economies to seek a new growth driver. This is where humanoid robots come in.

This is no longer a science fiction fantasy, but a realistic scenario supported by hard market data. According to a MarketsandMarkets report, the value of the global humanoid robot market is expected to grow from $2.03 billion in 2024 to $13.25 billion by 2029. This growth, with an annual CAGR of 45.5%, is impressive even to the most sceptical analysts. This momentum is driven by specific applications, ranging from healthcare and education to heavy industry, where precision and flexibility in work are becoming critical requirements.

Even more ambitious are the forecasts of Goldman Sachs Research, which estimate that the global humanoid market will reach $38 billion as early as 2035. Roland Berger’s forecasts present an even bolder vision. According to their analysis, universal humanoid robots could become widespread in 2030, generating up to $1.5 trillion in annual revenue by 2050 and reaching sales of 50 million units. These figures suggest that humanoids could become as integral to the economic landscape as production lines or autonomous mobile robots in warehouses are today.

However, any promise on this scale must be considered in the context of technological reality. Humanoid production remains costly and complex, and regulatory barriers and social acceptance could hinder their widespread adoption. Nevertheless, the first signs of a breakthrough are already visible: the costs associated with developing and implementing these machines have fallen by around 40% in the last year alone. This decline could pave the way for the technology to become more accessible and scalable.

Asia dominates the global market, accounting for almost 35% of humanoid robot revenue in 2023. Not only are China, India and Japan investing, they are also creating the institutional foundations for development – the Chinese government, for example, has set up a special robotics fund to support research and implementation. This signals that these countries are beginning to treat humanoids as part of their future economic infrastructure.

According to Interact Analysis, at least 16 companies worldwide are actively working on humanoids capable of moving on two legs or mobile bases and equipped with functional arms. These are not just theoretical projects, but actual prototypes that could become mass-market products in the coming years.

We are approaching a moment when we should ask ourselves, ‘When and to what extent will humanoids find their place in the economy?’ If technological innovation continues at its current pace and the regulatory environment remains favourable, humanoid robots could become the primary means of balancing the growing demands of the market with dwindling human resources. The story of the next industrial revolution — this time humanoid — will then begin to unfold before our eyes.

However, the road to the widespread adoption of humanoids will not be without barriers. Firstly, there is the cost. Although this has already fallen by 40% in the last year, it still poses a serious obstacle to widespread implementation. Companies must consider not only the purchase price, but also the costs of integration, maintenance, and system security. Secondly, there is the issue of technological complexity. Creating a machine that can move on two legs and adapt flexibly to a dynamic working environment requires advanced mechatronics, artificial intelligence algorithms, and reliable energy systems.

Legal regulation is also an important factor. The lack of globally consistent safety and liability standards could hinder market development. Who is responsible for an accident: the manufacturer, the operator, or perhaps the owner of the robot? These are questions that legislators must resolve before humanoids enter factories, hospitals and schools on a large scale. Additionally, the issue of social acceptance must be considered: are societies ready to trust machines that perform tasks and visually resemble humans?

Development scenarios could therefore diverge in several directions. In an optimistic scenario, favourable regulations, rapid cost reductions and further technological progress could lead to humanoids becoming commonplace in industry and services by 2030. In the more cautious scenario, legal and cost barriers will impede progress, limiting applications to specific sectors such as medicine and logistics. Finally, the pessimistic scenario assumes that a lack of public acceptance and unclear regulations will lead to market fragmentation and a slowdown in investment.

5 examples of humanoid robots

Supported by Hyundai, Atlas from Boston Dynamics sets new standards in movement precision. It is capable of rotating its limbs 180 degrees and operating in confined factory spaces. Tesla, on the other hand, has taken a bold step by adapting its Full Self-Driving technologies for use in the Optimus robot, sparking both fascination and controversy. Figure AI, which has received $675 million in funding, is developing humanoids to work on BMW production lines. The company is emphasising an ethical approach, collaborating with OpenAI and Microsoft. Apollo, developed in successive iterations over many years, can already carry 25-kilogram loads and may soon be working on the Mercedes production line. And Ameca? The most ‘human’ of the five, it can tell jokes thanks to GPT-4 and express emotions through its facial expressions.

Atlas: a pioneer in humanoid robotics

The story of Atlas is an example of how an experimental research project can be a turning point in industrial automation processes. Recognised as a pioneer in the field of humanoid robotics, Boston Dynamics has been stirring emotions for years, both within the scientific community and among engineers in the industry. In collaboration with Hyundai, the company is now entering a phase in which its bold concepts are gaining practical relevance. Once a technology demonstrator, Atlas is now becoming a tool ready for use in real industrial conditions.

Of particular interest is the fact that the latest Atlas model has moved away from hydraulic solutions in favour of a fully electric design. This is not just a change in the robot’s architecture, but a redefinition of standards. The precision, fluidity and dynamics with which this machine moves are reminiscent of scenes from science fiction films — except here we are dealing with reality, as can be seen in Boston Dynamics’ laboratories. The robot can rotate its body parts by 180 degrees, and its movements replicate human biomechanics so closely that, in some cases, they even surpass them thanks to the customised, highly flexible actuators used. This groundbreaking solution opens the way to tasks that would be impossible for humans.

It is worth noting that the engineers did not solely focus on spectacular skills. Reducing the robot’s turning radius is an important detail in an industrial environment. Production halls, assembly lines and warehouses are full of tight spaces where every extra centimetre of manoeuvring space translates into real time savings and minimised downtime. Atlas is becoming increasingly adept at working in these conditions, making it a practical addition to factory production lines.

The robot’s new ‘hands’ are equally important. Inspired by the earlier hydraulic model, they have been simplified to three fingers. While this may seem like a regression in terms of the number of components, in practice it is a step towards greater functionality and efficiency. In engineering, less often means more: the simplified hand design increases strength and reliability while enabling precise tasks that require stability and repeatability of grip rather than imitation of the human hand.

Atlas is no longer just a spectacular display of technology. It is an increasingly sophisticated tool that could transform our perception of machines working alongside humans. The essence of the automation revolution is not just about replacing humans; it is about creating solutions that increase the safety, efficiency and flexibility of processes.

Tesla Optimus: Elon Musk’s vision

When Elon Musk unveiled Tesla’s humanoid robot, aptly named Optimus, to the world, many commentators felt that this marked a turning point — not only for Tesla, but also for the field of industrial robotics. On the one hand, Optimus draws heavily on Tesla’s automotive technology; on the other, it offers a vision of the future in which work automation enters a whole new level. This is not just another robot for repetitive tasks. It is an experiment with the ambition to transform the relationship between humans and machines.

The robot was built using Full Self-Driving (FSD) software, which is found in Tesla cars. This software gives the humanoid the ability to make decisions in real time, understand its surroundings, and respond to changing and often unpredictable working conditions. In practice, this means that the robot can perform tasks with a level of precision that was previously thought to be the exclusive domain of human skill. Introducing a fusion of sensors and algorithms, which previously revolutionised the automotive industry, into the world of robotics is a strategic — and perhaps symbolic — move.

Safety, a key issue in both autonomous transport and robotics, has been prioritised here. Thanks to system redundancy, Optimus is more than just a prototype for media demonstrations. It is a design prepared for industrial scenarios where reliability is a prerequisite, not a luxury. In environments requiring continuous, repetitive work, it can be expected that this robot will not only meet the challenges, but also pave the way for the further automation of processes still awaiting breakthrough solutions.

However, it is impossible to talk about Optimus without mentioning the reaction to Elon Musk’s presentation. His comparison to the TV series Westworld fuelled emotions — for some, it was a warning against losing control over technology; for others, it was an exciting glimpse of the inevitable. This controversy is no accident. This is not the first time Tesla has balanced on the edge of innovation and provocation. Optimus seems to embody this philosophy, combining hard data and engineering with a narrative that captures the world’s attention.

So, is Tesla’s humanoid robot a genuine step towards the future of work or a manifesto for technological audacity? The answer is not clear-cut, but one thing seems certain: Optimus is not just another research and development project. Like Tesla’s electric cars, it is a symbol of ambition that can push the boundaries of what is possible. In this sense, it’s about more than just robots. It’s about redefining automation and prompting a question that every engineer and entrepreneur will have to address sooner or later: are we ready for a world in which humanoids become as integral to everyday life as cars are today?

Figure AI: assistant on BMW production lines

At first glance, the figures are impressive: $675 million in funding and a valuation of $2.6 billion. Until recently, such sums in the world of humanoid robotics would have seemed like science fiction. Yet Figure AI demonstrates that the concept of integrating human-like machines into industrial processes is maturing and gaining momentum. In a world where automation is an everyday occurrence rather than a novelty, Figure AI has emerged with ambitions that go far beyond simply speeding up processes – it aims to set a new standard for human-machine interaction.

Figure AI does not build robots for show. Its latest model, Figure 02, is a design that transcends existing technological limitations. Its bipedal mobility, advanced fourth-generation dexterity with 16 degrees of freedom, human-level strength and three times more computing power than its predecessor make it a tool ready for use in industrial conditions, not just in the laboratory. This transition from the experimental phase to practical implementation is what has made Figure the talk of the industry today.

The proof? Tests at the BMW factory in Spartanburg. Figure 02 humanoid robots successfully placed sheet metal components in holders there, which then became part of the chassis. This may seem simple, but every production engineer knows that this task requires absolute precision and repeatability. A mistake here is not just an error; it poses a real risk of stopping the production line. Figure has gained something much more valuable than another successful test: knowledge of how to integrate multi-tasking humanoid robots into existing production systems, and of how to communicate with machines in real industrial conditions.
Against this backdrop, the company’s approach to ethics is noteworthy. At a time when public opinion on the use of artificial intelligence in the military is becoming increasingly polarised, Figure has made its position clear: ‘non-military applications’ form the basis of its development strategy. Its cooperation with giants such as OpenAI and Microsoft strengthens its technological position and builds its image as a socially responsible partner. This may prove to be an invaluable long-term advantage.

However, the Figure 02 humanoid robot is more than just a promise of improved production efficiency. It also raises questions about the future of work and ergonomics. A machine capable of autonomous two-handed manipulation, complex grasping, and dynamic coordination is invaluable where humans struggle with monotony, accident risk, or physical strain. ‘Improving ergonomics and work safety’ is no longer just an advertising slogan, but a real direction of change in production plants.

Figure humanoid robots are entering factories today; tomorrow, they may become standard in logistics or retail. Could this be the start of an era in which robots will be as commonplace on the production floor as welding arms are today? Everything points to yes. Rather than joining the race, Figure AI is raising the stakes by showing that a humanoid robot can be more than just an impressive prototype — it can be a real industrial worker.

Whether we like it or not, the industry must answer the question: are we ready for the moment when humanoid robots become an integral part of economic processes rather than an addition? Figure 02’s answer is that it already is.

Apollo – the one who can build your Mercedes

Standing at 173 centimetres tall and weighing 72 kilograms, it is almost as tall as a human being. However, rather than a heart, its ‘chest’ conceals an electronic display. It sees through cameras hidden in its eyes and wears black, heavy gloves that conceal precision grippers instead of skin. This is not a character from a science fiction film, but the eighth generation of a humanoid robot designed to work in environments where humans should not, or cannot, always be present.

Every movement of this machine is the result of years of research. Apollo, as the robot has been named, can bend at the waist and knees and carry loads of up to 25 kilograms. There is more to this ability than just steel and aluminium. At the heart of the design are actuators – as many as 32 in each unit – which underwent forty iterations before reaching their current level of efficiency. This number best illustrates the engineers’ determination. An idea or a vision alone was not enough – perseverance was needed to create a machine that is both powerful and precise.

The eighth version of Apollo is a milestone, but there is no room for stagnation in this story. Before the end of the year, the world will see the ninth version of the robot. It will be slimmer, with no visible wires, and have a larger battery for a longer operating time. However, the most intriguing element will be the screen in its ‘mouth’ – a feature that enables the robot to express emotions through facial expressions. This is a technological and symbolic gesture. The machine is beginning to communicate with humans in the most natural language – that of expression.

Humanoid robots are increasingly becoming more than just engineering demonstrations. Apollo demonstrates that consistency and iterative development can push the boundaries of what is possible. Each subsequent version is not just a cosmetic improvement; it is also a testament to the fact that 21^(st)-century industry is looking for solutions capable of meeting production, logistics and task demands in extreme conditions.

Mercedes is currently testing these robots for undemanding, repetitive tasks: Mercedes tests humanoid robots for low-demand, repetitive tasks

Ameca: the world’s most advanced humanoid robot ready to meet people?

The humanoid robot created by Engineered Arts exemplifies engineering precision and a visionary approach to automation. Its design combines technology with anthropology, giving the machine a human face and expression while redefining the concept of human interaction with technology.

The most fascinating aspect of this humanoid is the 27 motors located in its head. These are not just dry numbers, though; each of these units has the ability to capture nuances such as raised eyebrows or a slight smile, and micro-expressions, which have formed the basis of interpersonal communication for centuries. Thanks to these motors, conversations with robots cease to be artificial and cold experiences and begin to resemble dialogues with human beings. Built-in cameras in the eyes complete this illusion: the humanoid not only ‘sees’ the world, but also reacts to it in a way that recipients intuitively recognise as natural.

However, the creators’ ambitions do not end there. Integration with large language models, including OpenAI’s GPT-4, enables the robot to understand questions and provide coherent answers. This is an important step because it shifts the burden of interaction from a mechanical exchange of messages to a dialogue. The robot can tell jokes and even imitate the voices of celebrities — something that, just a few years ago, was more common in software experiments than in physical machines in social spaces. This combination of artificial intelligence and mechanical engineering marks a new chapter in the history of humanoid robotics.

It is worth noting, however, that this humanoid cannot yet walk. Lack of mobility is one of the key barriers that engineers are currently trying to overcome. Once this problem is solved, robots will gain a new dimension: they will become interactive ‘actors’ not only in entertainment and education, but also in everyday life, potentially acting as companions, guides or assistants. Mobility would enable them to be present in real human environments, making their role even more practical.

The industry considers these designs from two perspectives: the ‘wow’ effect and long-term applications. What seems like a futuristic attraction today could become an integral part of our social and professional infrastructure tomorrow. History shows that solutions designed for entertainment often become the foundation for innovations of strategic importance. Will this also be the case here? The answer is still emerging, but one thing is clear: the humanoid robot from Engineered Arts is more than just a curiosity; it is a sign of the direction in which 21^(st)-century robotics is heading.

The construction of a humanoid robot, or a little bit about the technical side

At first glance, a humanoid robot appears to be a uniform structure, but in reality, it is a precisely designed combination of four modules. Each module has a strictly defined role, and their interaction creates a mechanism that only functions through synergy. There is no room for chance in this arrangement: engineers consciously design a structure in which the boundaries between modules also form the basis of its reliability. This approach demonstrates that the construction of humanoid robots is not merely a copy of nature, but rather an engineering balancing act where the whole is greater than the sum of its parts.

Sensory module

In the world of humanoids, the sensor module serves as both a sense and a form of technological intuition – determining not only whether a robot can ‘see’, but also whether it can understand the space in which it operates. Humans acquire over 80% of their knowledge through sight, which is why the vision system lies at the heart of perception and navigation in humanoid designs. Tesla Optimus is an example of this approach, being equipped with eight cameras, LiDAR, depth sensors and ultrasonic sensors. These allow it to build three-dimensional models of its surroundings. Thanks to its integration with Occupancy Networks, the robot can react to changes in its environment almost immediately, and it is precisely this speed of adaptation that will determine whether humanoids become real tools in industry or just spectacular prototypes.

Control module

In industrial laboratories, the ‘brain’ of a humanoid robot is no longer just a metaphor; it is a precisely designed control module in which CPUs, GPUs and AI accelerators work together to perform real-time calculations. However, it is artificial intelligence-based software, rather than silicon, that enables the machine to ‘think’ – analysing images, sounds and contextual data to make decisions in fractions of a second. Thanks to large language models such as GPT, the humanoid can not only plan its path and react to changing conditions, but also adapt its greeting to the tone of a person’s voice. Consequently, what seemed like science fiction a decade ago is becoming commonplace: robots are learning, adapting and optimising their actions, providing industry with partners that not only follow commands, but also understand their environment.

Executive / mechanical module

When we think of humanoid robots, we tend to imagine machines that move with human grace. However, behind this fluidity lies extremely complex engineering, at the heart of which is the executive module. This module converts digital commands into real actions, step by step and movement by movement, whether that be lifting an object or performing precise operations requiring surgical accuracy. Its structure combines technologies that, just a decade ago, were confined to laboratories: servo motors and hydraulic drives provide the energy for movement; harmonic reducers amplify torque and maintain precision; and encoders track every millimetre of the trajectory, enabling the system to correct errors immediately.

It is like an orchestra of mechanics and electronics, where each element plays a leading role because only through perfect coordination can the dynamic stresses of human walking, for example, be managed. Designers emphasise durability and reliability for good reason: components must withstand constant work cycles while maintaining full efficiency. At the same time, auxiliary systems ensure efficiency and minimise wear and tear, meaning the robot is not just a showcase prototype, but a genuine working partner. The result? A machine that performs tasks with mathematical precision in a surprisingly natural way.

Power supply module

The power supply module is the silent heart of the robot, without which no movement, camera view or arm gesture would be possible. It provides uninterrupted and flexible power, enabling the robot to walk and perform precise operations in conditions that would be extremely challenging for humans. Inside the module is a sophisticated ecosystem comprising a battery management system (BMS) that monitors and protects the cells from overload; inverters that convert raw energy into a form suitable for the various components; and power distribution units (PDUs) that decide in real time where to direct and limit additional watts. This seemingly invisible process determines whether a humanoid can lift a heavy structural element or maintain precise movement during complex tasks. The system is complemented by a cooling system that prevents temperature from affecting performance.

Thanks to this integration, energy is not only supplied, it is also managed with surgical precision, eliminating the risk of interruptions and providing a reliable foundation. This is why a properly designed power module is so much more than just a source of energy: it guarantees that a humanoid robot can work for longer and more stably, bringing not only strength to the industry, but also the intelligence of millisecond decisions about power allocation.

Summary

Clearly, humanoid robots are no longer just futuristic experiments in research laboratories; they are becoming active participants in the global economy. Boston Dynamics, Hanson Robotics and Engineered Arts are demonstrating the impressive capabilities of their designs and building a narrative in which humans and machines share workplaces and everyday spaces. These are not just devices with strictly defined functions; they are systems equipped with artificial intelligence, sensors and the ability to interact. They have the potential to transform our understanding of technological cooperation in the years to come.

Investment dynamics speak for themselves: China, Japan and the United States are competing not only in research and implementation, but also in shaping the global standard for the entire industry. This can be seen in manufacturing, where humanoids can boost efficiency; in healthcare, where they can assist medical staff; and in education, where they can facilitate entirely new approaches to learning. Elon Musk goes even further, suggesting that the market value of humanoids could exceed that of the most advanced technologies currently considered to be the pinnacle of innovation.

This raises a question that goes beyond purely technical considerations: If the boundary between biology and technology is becoming increasingly blurred, what role should humans play? Will humanoids simply be another tool for automation, or will they begin to redefine the meaning of work, relationships, and presence? The evolution of humanoid robotics is not only an engineering challenge, but also a test of our social maturity and readiness for new models of coexistence.

Viewed from an industrial perspective, it is hard not to notice that we are witnessing the birth of something that seemed like science fiction just a decade ago. Today, however, it is a real process, accelerated by global competition and billions of pounds of investment. This means that the question of what it means to be human in the age of humanoids is becoming a strategic challenge for engineers, entrepreneurs, and entire societies, as well as a philosophical dilemma. This story is beginning now, and we will feel its consequences for decades to come.