Chongqing - Humanoid robots have quickly entered the public’s consciousness this year, becoming a major technological trend. While they are often called "humanoid," many of these robots lack the expected human-like faces. So, why do so many humanoid robots forgo facial features?

A humanoid robot running. (Photo/Wu Jiaju)

Humanoid robots are designed to replicate human movement, but many of them lack one key feature: a human-like face. So why is this the case?

Wu Changzheng, President of MagicLab, explains that humans have facial features primarily for functions like vision and hearing. Robots, however, can achieve these capabilities using cameras and microphones. MagicLab’s Z1 robot, for instance, recently competed in a 100-meter race at a robotics event, showcasing impressive functional abilities despite lacking a recognizable face.

"While humans have two eyes, robots can be equipped with a triple-camera system that offers a wider field of view, which is often more practical. Designing facial features just to resemble humans could actually limit the robot's perceptual performance," Wu said. MagicLab's robots prioritize functionality over facial design to maximize their capabilities.

Because humanoid robots are still focused on perfecting their movement rather than social interaction, omitting facial features allows for more flexibility in design. For example, instead of having a face, visual systems can be distributed across the robot’s head. Multiple cameras placed on the top, rear, or other curved surfaces allow for 360-degree perception without the obstruction that facial structures might cause. This design also enables easier integration of additional sensors.

BoosterRobotics, whose T1 robot recently won the humanoid division of the Robot Soccer World Cup, shares this philosophy. Vice President Zhao Weichen emphasized that the primary focus is on perfecting movement—such as walking, grasping, and interacting—rather than simulating human expressions.

Zhao also pointed out a potential downside to highly realistic facial features. If these features fall short of fully replicating human expression, they could trigger the "Uncanny Valley" effect, causing discomfort and unease among human users.

A lifelike humanoid robot designed to resemble Li Bai, the greatest poet in Chinese history. (Photo/Wu Jiazu)

On the other end of the spectrum, some companies specialize in creating highly realistic humanoid robots, focusing on advanced bionic and facial simulation.

For example, EXrobots has developed humanoid robots that resemble historical figures like the renowned Chinese poet Li Bai and physicist Albert Einstein, complete with strikingly realistic faces.

Li Boyang, Co-founder and President of EXrobots, emphasizes that facial simulation is not just about aesthetics. "In human communication, language is only part of the story. Expressions and body language often carry greater meaning in many scenarios," he explained.

Li believes that lifelike humanoid robots are the ideal platform for emotional AI, enabling richer and more natural human-robot interactions. This makes them well-suited for applications in entertainment, companionship, education, customer service, senior care, and even mental health therapy—roles that robots without facial simulation are not equipped to fulfill.

Another company, Elephant Robotics, is also working on realistic androids, but it acknowledges the significant technical challenges involved. Business Director Zhou Yuqi explained that mechanics and biomimicry are currently separate engineering domains. "Adding synthetic skin or clothing can impact a robot’s flexibility. If only the head looks human while the body remains mechanical, the effect feels inconsistent," Zhou said.

A humanoid robot playing soccer. (Photo/Wu Jiazu)

From a technical perspective, faceless and facially realistic robots are designed with different objectives in mind, though the line between them is beginning to blur.

For instance, BoosterRobotics’ latest K1 model incorporates a dual-camera setup that mimics eyes, giving the robot a "cuter" appearance while maintaining a wide visual range and strong image recognition capabilities. Similarly, Fourier Intelligence’s newly launched GR-3 robot features large, expressive eyes and more pronounced facial features than previous models.

Human faces can convey subtle and complex emotions through intricate muscle movements, a capability that current technology has yet to fully replicate. Effectively merging mechanical flexibility with nuanced emotional expression remains a major challenge.

From an industry standpoint, movement-focused humanoids and bio-realistic androids are advancing in parallel, each following its own trajectory toward the same overarching goal: transforming how we live and work. The State Council of the People’s Republic of China recently released its "Opinions on Deepening the 'AI+' Action," advocating for the development of intelligent terminals—including robots—and calling for the creation of fully integrated smart interactive environments.

Looking ahead, with anticipated advances in flexible materials, bionic actuators, and artificial intelligence, the real question may no longer be whether humanoid robots should have faces, but how to seamlessly blend functionality and expressiveness in the robots of tomorrow.