In October 2025, US-based Richtech Robotics introduced Dex, the company's first mobile humanoid robot designed specifically for industrial applications. The announcement marks a strategic shift for the company, which has built its reputation primarily in hospitality service robotics since its founding in 2016.
Dex represents a departure from the bipedal humanoid designs that have dominated recent robotics headlines. Instead of legs, Dex operates on a wheeled autonomous mobile robot (AMR) platform. According to Casella, a company representative quoted in the announcement, this design choice was deliberate and practical. The rationale centers on operational endurance and stability — two areas where legged robots have historically faced challenges.
The company states that Dex's wheeled platform enables continuous operation for more than four hours per charge cycle. This compares favorably to bipedal systems that must constantly manage balance and often exhaust battery reserves more quickly. Dex can also carry heavier payloads than legged alternatives, navigate confined spaces with faster response times, and maintain what the company describes as "rock-solid stability" in human environments. Energy consumption is also lower, according to the company.
Training methodology is another distinguishing feature. Dex combines real-world data with Nvidia Isaac Sim, an open reference robotics simulation framework. The robot first learns tasks virtually at what the company describes as an exponential rate, then implements those learned behaviors into live industrial settings. After initial simulation-based training, Richtech refines Dex's performance using real-world operational data. This hybrid approach allows the robot to be trained quickly with new data and adapted for specialized workflows.
The strategic context is important. Richtech currently operates approximately 450 robot deployments, though the company has not disclosed how many of these existing deployments are suitable for a wheeled humanoid-like robot. The industry is still awaiting the first customer announcement for Dex. Richtech's business model is anchored in robot-as-a-service (RaaS), which means customers subscribe to the robots rather than purchasing them outright.
The company's stated goal is to scale its fleet to over 1,000 active deployments. This expansion includes transitioning from hospitality service into industrial settings with Dex. The company also points to a broader challenge facing the US robotics market: a shortage of high-quality, real-world operational data needed to train and scale physical AI systems. Richtech positions itself as addressing this gap through its data infrastructure and deployment model.
This launch comes amid a crowded and competitive humanoid robotics landscape. The source material also references other notable systems, including 1X Technologies' NEO Gamma — a 5-foot-6-inch humanoid with camera-infused eyes behind a dark shield on a human-shaped head, which some observers have compared to the helmets worn by French electronic music duo Daft Punk. Randy Howie, co-founder of New York Robotics, a non-profit fostering robotic development in the New York area, has suggested that NEO Gamma is closer to entering the home than any other humanoid currently in development.
Agility Robotics' Digit V4 is also mentioned, billed as the "world's first commercially deployed humanoid robot." The system contains approximately 5,000 parts. Tim Smith, a spokesperson for Agility, told The Post that Digit 4 is the only humanoid robot currently working in warehouses and factories today.
Richtech has also been active in expanding human-robot interaction channels. The company launched a round-the-clock interactive livestream platform featuring its ADAM robot, allowing audiences worldwide to interact directly with the robot in real time. Wayne Huang, CEO of Richtech Robotics, framed this initiative as opening "a new chapter in human-robot interaction" by providing a global opportunity to communicate with embodied AI in a live, highly interactive setting.
Why it matters for European robot service
For European operators and service providers, the Dex announcement carries several implications worth examining.
First, the wheeled-versus-legged design debate has direct consequences for deployment economics. European manufacturing facilities, logistics hubs, and industrial sites often have infrastructure built around wheeled transport — floors designed for forklifts, pallet jacks, and automated guided vehicles. A wheeled humanoid can integrate into these environments without requiring significant facility modifications. The company's claims about longer battery life and lower energy consumption also matter in a European context where energy costs remain a significant operational concern for manufacturers.
Second, the simulation-to-reality training pipeline addresses a bottleneck that European robotics adopters frequently encounter. The shortage of high-quality, real-world operational data is not unique to the US market. European manufacturers considering physical AI systems face similar challenges in sourcing the data needed to train robots for specialized workflows. The approach of learning tasks virtually in simulation, then refining with real-world data, could reduce the deployment friction that has historically slowed robotics adoption in smaller and mid-sized European manufacturing operations.
Third, the robot-as-a-service model is particularly relevant for European buyers. Capital expenditure constraints are common across European manufacturing, especially among small and medium-sized enterprises. RaaS arrangements shift robotics from a capital purchase to an operational expense, which can lower the barrier to entry. The model also aligns with the service-oriented approach that many European industrial automation providers have adopted.
Fourth, the transition from hospitality to industrial settings is a pattern European service providers should monitor. Richtech's existing deployments are primarily in hospitality, but the company's stated ambition to scale to over 1,000 deployments includes a significant industrial component. This suggests that the company sees industrial applications as the growth market, and European industrial operators may eventually see Dex or similar systems offered through local service partners.
Fifth, the competitive landscape matters. The source material references other humanoid systems — 1X's NEO Gamma and Agility's Digit V4 — that are also vying for commercial traction. European buyers evaluating humanoid robots will have multiple options to consider, and the differences in design philosophy (wheels versus legs), training methodology, and business model will be important differentiators. The fact that Agility claims Digit 4 is the only humanoid currently working in warehouses and factories suggests that commercial deployments are still nascent, and the market remains open for systems that can demonstrate reliable, sustained operation.
Sixth, the livestream initiative with ADAM signals a broader trend toward remote interaction and telepresence in robotics. For European service providers, this could open new possibilities for remote monitoring, training, and customer engagement. The ability to interact with a robot in real time from anywhere in the world has implications for how robots are commissioned, maintained, and operated across distributed sites.
What buyers and operators should know
For organizations considering Dex or similar mobile humanoid systems, several practical points emerge from the source material.
**Deployment readiness is still unproven.** The company has not announced a first customer for Dex. While Richtech has approximately 450 existing deployments, the source material explicitly notes that it is unclear how many of these are ready for a wheeled humanoid-like robot. Buyers should treat claims about industrial readiness with appropriate caution until real-world deployments are documented and operational data is available.
**Battery life and payload specifications are company-provided.** The four-plus hours of continuous operation and heavier payload capacity are claims made by Richtech, not independently verified figures. Buyers should request detailed specifications and, ideally, reference deployments before making procurement decisions. The comparison to legged alternatives is also company-framed; independent benchmarking data is not provided in the source material.
**Simulation training is a differentiator, but validation is ongoing.** The combination of Nvidia Isaac Sim for virtual training followed by real-world refinement is an established approach in robotics, but the effectiveness depends on the quality of the simulation environment and the real-world data used for fine-tuning. The company's acknowledgment of a shortage of high-quality, real-world operational data in the US market is notable — it suggests that data acquisition remains a challenge even for companies building data infrastructure.
**RaaS pricing is not disclosed.** The source material confirms that Richtech's business model is anchored in robot-as-a-service deployments, but no pricing information, contract terms, or service-level agreements are provided. Buyers should not assume standard RaaS terms; each deployment will likely require individualized negotiation.
**The wheeled design has trade-offs.** While wheels offer advantages in battery life, payload, and stability, they also limit the robot's ability to navigate stairs, uneven terrain, or environments designed for human ambulation. Buyers should assess whether their facilities present obstacles that a wheeled platform cannot handle. The company's claim about navigating tight spaces is relevant, but "tight" is not defined with specific dimensions.
**Fleet scaling ambitions are stated, not achieved.** The goal of scaling to over 1,000 active deployments is an aspiration, not a current reality. The company has approximately 450 deployments today, and the transition from hospitality to industrial settings is still in its early stages. Buyers should evaluate the company's track record in hospitality as a proxy for its ability to deliver on industrial commitments.
**The competitive field is active.** The source material references multiple humanoid robots in development or early deployment. Buyers should compare systems across dimensions that matter for their specific use cases: battery life, payload capacity, navigation capabilities, training requirements, and total cost of ownership. The fact that Agility claims its Digit 4 is the only humanoid currently working in warehouses and factories is a significant data point — it suggests that commercial humanoid deployments are still extremely rare.
**Data infrastructure matters.** Richtech emphasizes its focus on developing robotics and the data infrastructure to make its systems more intelligent. For buyers, this raises questions about data ownership, data security, and how the data generated by their robots will be used. These are contractual issues that should be addressed before signing any agreement.
**The livestream initiative is a novel engagement model.** The round-the-clock ADAM livestream allows audiences to interact with the robot in real time. While this is primarily a marketing and public engagement initiative, it also demonstrates the company's confidence in its remote interaction capabilities. For industrial buyers, this could suggest that remote monitoring and interaction features are mature enough for public demonstration.
**Timeline expectations should be conservative.** The source material does not provide specific delivery timelines, production volumes, or availability dates for Dex. The industry is still waiting for the first customer announcement. Buyers should expect that early deployments will be pilots or proof-of-concept projects rather than large-scale rollouts.
**European-specific considerations are not addressed.** The source material does not mention European availability, certifications (such as CE marking), or local support infrastructure. European buyers will need to verify these details directly with the company or through local distributors.
In summary, Dex represents a notable entry in the mobile humanoid category, with a pragmatic wheeled design and a training approach that leverages simulation to accelerate learning. However, the lack of announced customers, undisclosed pricing, and unverified performance claims mean that buyers should approach with measured expectations. The robot-as-a-service model may lower entry barriers, but the details of such arrangements remain to be negotiated on a case-by-case basis.
Sources
Richtech Robotics unveils its first humanoid robot for ‘real-world work’
Published by Vigla Media OÜ (Estonia).