The humanoid robotics sector is entering a phase where demonstration circuits, trade-show appearances, and third-party technical evaluations are becoming as important as the hardware itself. The upcoming Robotics Summit & Expo serves as the latest venue where Unitree, the Chinese robotics manufacturer, will present its G1 humanoid platform. This is not the first time the G1 has been placed before an international audience; the robot has already appeared at the Humanoids Summit in Tokyo and at Automation World in Seoul, where it was shown alongside other Chinese humanoid manufacturers. What is notable about these appearances is the pattern they reveal: Chinese humanoid makers are systematically using global expositions to build visibility, test market reactions, and position themselves for commercial deployment beyond their home market.
The context for this demonstration extends beyond a single trade show. Fraunhofer IPA, the German research institute known for applied production technology, has taken the Unitree G1 into its evaluation facilities. The institute tested the robot’s power consumption and assessed its suitability for production environments. This is a significant development because it moves the conversation about humanoids from speculative marketing to measurable engineering data. The G1, in its EDU-4 configuration with Dex3-1 three-finger hands, delivered in May 2025 with firmware version 1.04, was subjected to a defined standard scenario. The results, published by Fraunhofer IPA, provide a rare glimpse into the actual energy demands of a commercial humanoid platform.
The broader context also includes a wave of Chinese companies—AGIBOT, Fourier, Leju, Huawei, and Unitree—confirming participation in the Smart Factory & Automation World (AW) 2026, which describes itself as “Asia’s leading manufacturing AX expo.” These companies will exhibit and speak at the China Humanoid Robot Conference, branded as “China Humanoid: First Journey to Korea,” affiliated with AW 2026. The conference will feature keynote addresses from Jiang Chengyi, head of solutions at Unitree; Zhou Bin, co-founder of Fourier; and Ren Guangjie, head of solutions at Leju. Their topics will cover China’s humanoid robotics strategies, commercialization roadmaps, and global expansion plans. This clustering of Chinese humanoid makers at international events suggests a coordinated push to establish a foothold in overseas markets, including Europe.
Key findings
The source material yields several concrete findings that merit attention from industry observers.
First, Unitree’s G1 has been demonstrated at multiple international venues. The Humanoids Summit in Tokyo, which took place on May 29, featured the G1 in a public setting. The same robot was also shown at Automation World in Seoul, where it was highlighted alongside other Chinese humanoid makers. These are not isolated appearances; they form a pattern of sustained international exposure. The G1 has also been brought to the Humanoids Summit in Mountain View, California, where it was presented by Ultimate Fighting Bots, a company that appears to use the platform for entertainment or combat-oriented demonstrations. At that event, a conference attendee was photographed touching the G1, indicating that the robot is designed for close interaction with the public.
Second, the Fraunhofer IPA evaluation provides hard numbers on the G1’s power consumption. In stationary operation, the robot consumes approximately 154 watts on average. When walking on a flat surface, consumption rises to approximately 272 watts. Walking on a 10 percent incline increases consumption to approximately 283 watts. For the defined one-hour standard scenario, the average power consumption is approximately 239 watts. These figures are significant because they give potential industrial buyers a baseline for energy costs and battery life planning. The Fraunhofer IPA evaluation also revealed limitations in the technology, although the source material does not specify the exact nature of those limitations. What is clear is that the institute learned a great deal about the current constraints of humanoid robots when applied to production environments.
Third, the competitive landscape is crowded and increasingly international. AGIBOT claims to be the world’s No. 1 company in global robot shipments and market share. The company plans to demonstrate its X2 and G2 humanoids at the Robotics Summit & Expo. Leju will show its Kuavo 4th Generation Pro and Kuavo 4 Pro robots. Unitree will show its G1. Huawei, a leading player in China’s core technology development, will present real-world applications of its embedded humanoid robotics technologies. This lineup suggests that the Robotics Summit & Expo will serve as a showcase for Chinese humanoid capabilities, with multiple companies vying for attention from potential buyers, integrators, and investors.
Fourth, there is a notable human element to the humanoid robotics story. Zach Vinegar, cofounder and CTO of Gatlin Robotics, and Isaac Qureshi, cofounder and CEO of the same company, took their Unitree G1 humanoid for a spin at the Humanoids Summit. Gatlin Robotics is a cleaning service humanoid company, indicating that the G1 is being used as a development platform for service applications. The source material also notes that there are concerns that robots will take people’s jobs and invade their privacy. Bot builders, however, say their products are designed to help humans, not replace them. This tension between public anxiety and industry messaging is a recurring theme in the humanoid sector, and it is likely to shape regulatory and adoption dynamics in Europe.
Fifth, the global expansion of Chinese humanoid makers is not limited to trade shows. The source material includes an image caption noting that Unitree’s G1 humanoid robots were shown during a martial arts demonstration, and another image shows a Tesla Optimus robot handing out candy in front of the Nasdaq MarketSite in New York in 2025. These images, while not central to the analytical content, illustrate the range of public-facing demonstrations that humanoid robots are undergoing. The Japan Times headline “China sends robots out into the world to learn” captures the strategic intent behind this international exposure: Chinese humanoid makers are using global appearances to gather data, refine their products, and build relationships.
What it means for European operators
For European operators considering the adoption of humanoid robots, the source material offers several actionable insights, as well as a number of open questions that the industry has not yet answered.
The Fraunhofer IPA power consumption data is the most concrete takeaway. A humanoid robot that consumes approximately 154 watts in stationary operation and approximately 272 watts when walking on a flat surface is not an energy-intensive machine by industrial standards. For comparison, a typical industrial robot arm can draw several kilowatts during operation. The G1’s average consumption of approximately 239 watts over a one-hour standard scenario suggests that energy costs will not be the primary barrier to adoption. However, the battery life implications are not disclosed in the source material. European operators will need to know how long the G1 can operate on a single charge in a real production environment, and the source material does not provide that figure. It is also not disclosed how long recharging takes, what the battery degradation curve looks like, or what the total cost of ownership would be over a multi-year deployment. These are material gaps that Fraunhofer IPA’s evaluation may address in future publications, but they are not available in the current source material.
The suitability of the G1 for production environments is another area where the source material is partially informative. Fraunhofer IPA tested the robot and learned a great deal about the current limitations of the technology. The source material does not specify what those limitations are, which is a significant omission for European operators who are evaluating the G1 for factory floor use. It is reasonable to infer that the limitations relate to payload capacity, endurance, precision, or software robustness, but the source material does not confirm any of these. European operators should therefore treat the G1 as a promising but not yet fully validated platform for production environments. The fact that Fraunhofer IPA is ready to test future humanoid robots suggests that the institute sees value in continued evaluation, but it also implies that current generations of humanoids have not yet met all the requirements for widespread industrial deployment.
The competitive landscape is another factor that European operators should consider. AGIBOT claims to be the world’s No. 1 company in global robot shipments and market share. If this claim is accurate, it suggests that the humanoid market is already consolidating around a few Chinese players. Unitree, Fourier, Leju, and Huawei are all actively pursuing commercialization and global expansion. This means that European operators will have multiple options when selecting a humanoid platform, but it also means that the market is still in flux. Standards, interfaces, and software ecosystems are not yet mature, and European operators may face integration challenges if they commit to a platform that does not become a market standard.
The public perception issue is also relevant for European operators. The source material notes that there are concerns that robots will take people’s jobs and invade their privacy. Bot builders say their products are designed to help humans, not replace them. This is a familiar narrative in the robotics industry, but it carries particular weight in Europe, where labor unions, works councils, and data protection regulations are more established than in many other regions. European operators who deploy humanoid robots will need to manage these concerns proactively, through transparent communication, worker retraining programs, and clear policies on data collection and privacy. The source material does not provide guidance on how to do this, but it does confirm that the concerns are real and widely reported.
The international demonstration circuit also has implications for European operators. The fact that Unitree’s G1 has been shown at the Humanoids Summit in Tokyo, Automation World in Seoul, and the Humanoids Summit in Mountain View, California, suggests that the company is serious about global expansion. The G1 has also been used by Gatlin Robotics, a cleaning service humanoid company, indicating that the platform is being adapted for service applications beyond industrial settings. This is relevant for European operators because it suggests that the G1 is not just a research platform; it is being commercialized for real-world use. However, the source material does not disclose pricing, delivery lead times, warranty terms, or service and support arrangements for the European market. These are critical factors for any capital equipment purchase, and their absence from the source material means that European operators will need to seek this information directly from Unitree or its local partners.
The upcoming Smart Factory & Automation World (AW) 2026, described as “Asia’s leading manufacturing AX expo,” will feature AGIBOT, Fourier, Huawei, Leju, and Unitree. The affiliated China Humanoid Robot Conference, branded as “China Humanoid: First Journey to Korea,” will include keynote addresses from executives at Unitree, Fourier, and Leju. Their topics will cover China’s humanoid robotics strategies, commercialization roadmaps, and global expansion plans. For European operators, this conference represents an opportunity to hear directly from the companies that are shaping the humanoid market. It also signals that Chinese humanoid makers are looking beyond their domestic market and are actively courting international customers. Whether this will translate into a strong European presence—with local service centers, spare parts inventories, and regulatory compliance—remains to be seen. The source material does not disclose any specific plans for European operations.
The source material also references a Tesla Optimus robot handing out candy in front of the Nasdaq MarketSite in New York in 2025. This is a reminder that the humanoid race is not limited to Chinese companies. Tesla, a US-based company, is also pushing its humanoid platform into public view. For European operators, this means that the humanoid market will be contested by multiple global players, each with different strengths and weaknesses. The source material does not provide a direct comparison between Tesla Optimus and Unitree G1, so European operators should not draw conclusions about relative performance from this material alone.
In summary, the source material provides a snapshot of a humanoid robotics industry that is moving from demonstration to commercialization. Unitree’s G1 has been shown at multiple international venues, evaluated by Fraunhofer IPA, and used by at least one service robotics company. The power consumption data from Fraunhofer IPA is concrete and useful, but the source material does not disclose the limitations that the institute identified, nor does it provide pricing, delivery, or support information for the European market. European operators should approach the G1 and other Chinese humanoid platforms with cautious optimism, recognizing that the technology is advancing rapidly but that critical commercial details remain undisclosed. The upcoming Robotics Summit & Expo and the AW 2026 conference in Korea will provide additional opportunities for European operators to evaluate the technology and engage with the manufacturers directly. Until then, the prudent approach is to treat humanoid robots as a promising but not yet fully proven technology for European production and service environments.
Sources
Published by Vigla Media OÜ (Estonia).