Humanoid Robot Builders

A cheatsheet exploring major companies developing advanced humanoid robots, their philosophies, key robots, funding, recent developments, and target applications.

Last Updated: May 2025

Boston Dynamics

Key Information
  • Founded: 1992 (spun off from MIT). Acquired by Hyundai Motor Group in 2021.
  • Headquarters: Waltham, Massachusetts, USA.
  • Key Figures: Robert Playter (CEO). Marc Raibert (Founder and Chairman).
  • Flagship Humanoid: Atlas (new all-electric version unveiled April 2024).
  • Other Key Products: Spot (quadruped), Stretch (logistics robot).
  • Parent Company: Hyundai Motor Group.
  • Official Website: bostondynamics.com
Origin & Founding Vision

Boston Dynamics originated as a spin-off from the Massachusetts Institute of Technology (MIT) in 1992. For decades, it has been a pioneer in developing highly mobile robots, initially with significant funding from DARPA. Their vision has consistently focused on pushing the boundaries of robot locomotion and manipulation in complex environments. The acquisition by Hyundai in 2021 aims to leverage Boston Dynamics' technology for broader industrial and potentially consumer applications.

Key Details
  • Early Focus: Developing robots that could maneuver in challenging, real-world terrains, often for defense and research applications.
  • DARPA Influence: Early versions of Atlas were developed with funding and oversight from the Defense Advanced Research Projects Agency (DARPA).
  • Evolution: Over the years, the company has transitioned towards commercial applications while continuing to advance the fundamental capabilities of its robots. The new electric Atlas is a testament to this, moving away from hydraulics for commercial viability.
Philosophy & Approach

Boston Dynamics' philosophy centers on creating robots with advanced mobility, dexterity, and intelligence that can tackle tasks in unstructured or dynamic environments. They emphasize dynamic balance, agile movement, and sophisticated perception to enable robots to operate effectively alongside humans or in hazardous conditions. Their design approach often prioritizes functionality and efficiency of movement over strict human biomimicry for specific tasks, aiming for robots to function effectively in human-centric environments.

Core Tenets
  • Dynamic Locomotion: A hallmark of their robots, enabling them to traverse complex terrains, maintain balance when disturbed, and perform agile maneuvers.
  • Purposeful Movement Design: Movements are optimized for task execution, efficiency, and effectiveness, which may sometimes appear unconventional compared to human motion. The electric Atlas is capable of all the agility its hydraulic predecessor had.
  • Real-World Utility: Focus on developing robots that can perform useful work in real-world settings, moving beyond lab demonstrations.
  • Iterative Development: Continuous refinement of hardware and software based on real-world testing and feedback. The new electric Atlas is a testament to this, moving away from hydraulics for commercial viability.
Leadership

Robert Playter serves as the Chief Executive Officer (CEO) of Boston Dynamics. Marc Raibert, the founder, is the Chairman. Aaron Saunders is the Chief Technology Officer (CTO).

Key Robots & Technologies

The flagship humanoid is Atlas, with a new all-electric version unveiled in April 2024, designed to be stronger, more agile, and with an enhanced range of motion. Other well-known robots include Spot (a versatile quadruped) and Stretch (a mobile robot for warehouse case handling).

Atlas (Electric Version)
  • Type: All-electric humanoid robot.
  • Capabilities: Designed to be stronger and more agile than its hydraulic predecessors, with a greater range of motion. Capable of lifting heavy items, managing awkward loads, and performing complex tasks like sorting car parts. Features exceptional balance and agility with dynamic locomotion, including running, crawling, and cartwheeling, enabled by custom-designed motors.
  • Technology: Utilizes machine learning for vision. A partnership with LG Innotek aims to install next-generation vision sensing components to improve perception, especially in low visibility or dark environments. Operates with pre-programmed logic and real-time sensor data for autonomous decision-making. Shared reinforcement learning training pipeline with the Robotics & AI Institute to build dynamic and generalizable mobile manipulation behavior.
  • Design: Humanoid form factor to navigate environments built for humans, though its movements are optimized for efficiency and may transcend typical human motions. Height: Approximately 1.5 m; Weight: Around 75 kg.
Other Robots
  • Spot: A highly mobile and customizable quadruped robot used for inspection, data collection, public safety, and remote operation in various industries.
  • Stretch: A mobile robot designed for truck unloading and palletizing in warehouses and distribution centers.
Long-Term Vision / Role of Humanoids

Boston Dynamics aims to create robots that can go where people go and do what people do, but also extend beyond human limitations in terms of strength, endurance, and operation in hazardous environments. The focus is on practical applications that solve real-world labor challenges and improve safety and efficiency.

Funding & Resources

Acquired by Hyundai Motor Group in June 2021 for a reported $1.1 billion. This provides significant financial backing and resources. Hyundai plans to purchase "tens of thousands" of Boston Dynamics robots in the coming years and will integrate its manufacturing capabilities to help Boston Dynamics grow.

Recent Developments (2024-2025)

The new all-electric Atlas, unveiled in April 2024, is planned for deployment in Hyundai's U.S. factory (Metaplant in Georgia) later in 2025 for real-world testing and task optimization. In May 2025, Boston Dynamics announced a partnership with LG Innotek to integrate advanced vision sensing components into Atlas. They also formed a partnership with the Robotics & AI Institute for reinforcement learning for Atlas. Continued focus on enhancing Atlas's autonomous capabilities and practical task performance, with collaborations with NVIDIA, Google DeepMind, and Toyota Research Institute.

Key Announcements
  • Electric Atlas Launch (April 2024): Introduction of the significantly redesigned, all-electric Atlas, showcasing improved strength, agility, and a sleeker form factor aimed at commercial applications.
  • Hyundai Factory Deployment (Late 2025): Plans to test and refine the new Atlas in Hyundai's automotive manufacturing plant in Georgia, USA.
  • LG Innotek Partnership (May 2025): Collaboration to install next-generation vision sensing components into Atlas for improved perception.
  • Robotics & AI Institute Partnership (May 2025): Collaboration to establish a shared reinforcement learning training pipeline for the new electric Atlas.
  • Enhanced Autonomy & Collaborations: Ongoing development of Atlas's AI, enabling it to perform complex tasks with greater independence, supported by partnerships with NVIDIA, Google DeepMind, and Toyota Research Institute.
Target Audience & Use Cases

Initially, manufacturing and logistics sectors, particularly automotive with Hyundai. Atlas is engineered to outperform human workers in strength and reliability, capable of lifting heavy items and managing awkward loads. Future applications could extend to construction, R&D, industrial inspections, search and rescue, and potentially hazardous environments, leveraging Atlas's advanced mobility and manipulation capabilities.

Tesla

Key Information
  • Project Announced: August 2021 (Tesla AI Day).
  • Headquarters: Austin, Texas, USA (Tesla, Inc.).
  • Key Figures: Elon Musk (CEO of Tesla).
  • Flagship Humanoid: Optimus (also referred to as Tesla Bot). Currently demonstrating Gen 2, with Optimus Gen 3 anticipated with design updates and potential sales starting late 2025 or 2026.
  • Parent Company: Tesla, Inc.
  • Official Website: tesla.com/ai
Origin & Founding Vision

Tesla announced its humanoid robot project, Optimus, at its AI Day in 2021. The vision, as articulated by Elon Musk, is to create a general-purpose humanoid robot capable of performing unsafe, repetitive, or boring tasks, eventually addressing labor shortages and transforming the nature of work. Musk believes Optimus could become more valuable than Tesla's car business.

Philosophy & Approach

Tesla's approach leverages its existing expertise in AI (particularly from its Full Self-Driving program), battery technology (like the 4680 battery pack in Gen 3), motors, electronics, and advanced manufacturing. The philosophy is to develop a highly capable yet eventually affordable humanoid robot (potentially $10,000-$20,000) that can be mass-produced and integrated into various industries and potentially homes. Emphasis is placed on real-world AI, vision-based navigation and manipulation, and efficient inference hardware.

Core Tenets
  • AI-Driven Autonomy: Utilizing Tesla's advanced AI for vision and planning to enable Optimus to navigate and interact with the physical world, with the aim for Optimus to automatically understand user habits and routines.
  • Scalable Manufacturing: Designing Optimus for mass production, aiming to make it affordable over time. Tesla aims to produce over 1,000 units for internal use in 2025 and scale to 50,000 in 2026.
  • Real-World Task Performance: Focusing on tasks that are "boring, repetitive, and dangerous" to provide practical value in factories, homes, hospitals, and disaster zones.
  • Iterative Development: Rapidly developing new generations (Gen 1, Gen 2, anticipated Gen 3) with significant improvements in each iteration.
  • Integrated Design: Developing key components like actuators, AI chips, motors, batteries, and power electronics in-house.
Leadership

The Optimus project is spearheaded by Tesla CEO Elon Musk, who provides the overarching vision and direction. Various engineering teams within Tesla are responsible for AI, hardware, and software development.

Key Robots & Technologies

Optimus Gen 2 is a frequently demonstrated version, featuring Tesla-designed actuators, improved balance, faster walking speed (target 2.2 m/s, current 1.8 m/s), and more dexterous hands (11-DoF, with 22-DoF anticipated for production units). It leverages Tesla's Full Self-Driving (FSD) computer and neural network technology. An Optimus Gen 3 is anticipated with significant design updates, including new arm and ankle geometry, potentially lighter materials (using 4680 battery pack made Gen 3 22 lbs lighter), and improved efficiency.

Optimus Gen 2 / Anticipated Gen 3
  • Height & Weight (Gen 2): Approx. 5'8” (1.73 m), Around 125 lbs (57 kg). Gen 3 is reportedly 22 lbs lighter due to the 4680 battery pack.
  • Mobility: Improved balance, faster walking speed (Gen 2 target 2.2 m/s, current 1.8 m/s; Gen 3 aims for smoother descent control). Recent videos show improved dance flexibility and control.
  • Manipulation: Gen 2 hands have 11-DoF with tactile sensing; production units and Gen 3 aim for 22-DoF hands, capable of delicate tasks like handling eggs or tying shoelaces. Actuators for hands may be moved into forearms in Gen 3.
  • AI & Perception: Utilizes Tesla's FSD computer, vision-based neural networks, and Dojo supercomputer for training. Aims to understand user habits and daily routines.
  • Actuators: Custom Tesla-designed actuators. Gen 3 may feature a new linear mechanical system operating under 40 dB.
  • Battery: Gen 2 has a 2.3 kWh battery pack; Gen 3 incorporates Tesla's 4680 battery pack for lighter weight and improved balance/efficiency.
Long-Term Vision / Role of Humanoids

Elon Musk envisions Optimus becoming more valuable than Tesla's car business, potentially transforming the global economy by creating an "age of abundance" where physical labor is optional and goods/services are cheaper. The long-term goal is a general-purpose robot that can perform a wide array of tasks in various environments including homes, factories, hospitals, and disaster zones.

Funding & Resources

The Optimus project is internally funded by Tesla, Inc., leveraging the company's substantial R&D budget and resources in AI, manufacturing, and battery technology.

Recent Developments (2024-2025)

Ongoing demonstrations of Optimus performing tasks like folding laundry, sorting objects, and dancing with improved flexibility. Limited production of Optimus reportedly started at Tesla's Fremont Factory in Q1 2025, with plans for over 1,000 units for internal use in 2025. External sales are projected for 2026. Optimus Gen 3 is anticipated with significant design updates and potential sales starting late 2025. Leaked images suggest a redesigned head, neck, and shoulders for Gen 3.

Target Audience & Use Cases

Initial deployment in Tesla's own manufacturing facilities for tasks like material handling and assembly, potentially playing a key role in Cybercab production. Future applications are envisioned in broader manufacturing, logistics, healthcare, disaster response, and eventually household assistance and companionship. The target price is aimed to be relatively affordable, potentially $10,000-$20,000 in the long term.

Figure AI

Key Information
  • Founded: 2022.
  • Headquarters: Sunnyvale, California, USA.
  • Key Figures: Brett Adcock (Founder and CEO).
  • Flagship Robots: Figure 01, Figure 02.
  • Valuation: ~$2.6 billion (as of early 2024). Potentially seeking $1.5B at a $39.5B valuation in early 2025.
  • Official Website: figure.ai
Origin & Founding Vision

Founded in 2022 by Brett Adcock, Figure AI aims to develop general-purpose humanoid robots to address labor shortages and perform dangerous jobs, initially targeting logistics and manufacturing. Their long-term vision is to make labor optional, seeing a $40 trillion annual market for human labor.

Philosophy & Approach

Figure AI emphasizes rapid development, prototyping, and testing to bring humanoids to market quickly. Their philosophy centers on a human-centric design, enabling robots to learn from human demonstrations and integrate safely into human environments. A key belief is that tight integration between hardware and AI software (like their AI system Helix and integration with OpenAI) is critical for performance. Safety is a core tenet.

Core Tenets
  • Speed and Agility: Prioritizing rapid prototyping and iteration to accelerate development.
  • Human-Centric Design: Robots are designed to look like humans to facilitate data gathering from human demonstrations and ensure safer integration into human workspaces.
  • Integrated AI and Hardware: Believes that AI software integration is a key differentiator and cannot be outsourced. Focus on real-time AI systems for physical robots.
  • Safety First: Ensuring robots can interact safely with humans in the workplace is essential.
  • Real-World Deployment: Aiming for practical deployment in commercial labor markets before consumer markets, with Figure 02 being tested in BMW's manufacturing facilities.
Leadership

Led by Founder and CEO Brett Adcock, an entrepreneur with a track record of building and exiting companies.

Key Robots & Technologies

Figure 01 was their initial prototype. Figure 02 is their more advanced humanoid, designed for industrial use with enhanced mobility, dexterity, and AI capabilities, including highly flexible, human-like fingers. Figure's robots are powered by their in-house AI (Helix) and manufactured at their BotQ facility. They have integrated OpenAI's technology for voice and reasoning capabilities.

Figure 02
  • Height & Weight: 5'6" (168 cm), 154 lbs (70 kg) - *Note: some sources slightly vary, this is from original doc.
  • Payload & Speed: Payload capacity of 44 lbs (20 kg), top speed of 2.7 mph - *Note: from original doc.
  • Endurance: Up to 5 hours on a single charge - *Note: from original doc.
  • Manipulation: 16 degrees of freedom in its hands, allowing for complex tasks like opening doors, climbing stairs, and handling tools. Features highly flexible, human-like fingers.
  • Perception & AI: Equipped with RGB cameras and powered by NVIDIA RTX GPUs. Integrated OpenAI models for language understanding and reasoning. Figure's own AI system, Helix, drives the robot's actions.
  • Design: Focus on versatility, collaboration with humans, and adaptability to existing industrial ecosystems.
Long-Term Vision / Role of Humanoids

Figure AI's long-term goal is ambitious: to deploy billions of humanoids across the globe to perform most manual labor, effectively making labor optional and creating a better life for future generations. They see humanoids as a general-purpose solution for a world designed for the human form, aiming to produce 100,000 robots eventually.

Funding & Investors

Raised $70 million in a Series A round in May 2023. In early 2024, secured around $675 million in a funding round including investments from Jeff Bezos (through Physical Intelligence), Microsoft, Nvidia, OpenAI, and Intel Capital, valuing the company at approximately $2.6 billion. Reports in early 2025 suggested Figure AI was eyeing an additional $1.5 billion in funding at a potential valuation of $39.5 billion. Has partnerships with OpenAI and BMW for AI development and deployment testing.

Recent Developments (2024-2025)

Announced a partnership with BMW in January 2024 to deploy Figure humanoids in their Spartanburg, South Carolina manufacturing facility. As of March 2025, a single Figure 02 robot was operating at the plant, performing tasks like retrieving metal sheet parts. BMW plans to share more details in May 2025. Figure AI became revenue-generating in late 2024 with this deployment. Collaboration with OpenAI continues for next-generation AI models. Scrutiny arose in April 2025 regarding the scale of the BMW deployment versus initial claims.

Target Audience & Use Cases

Initially targeting labor-intensive tasks in manufacturing (e.g., automotive with BMW), logistics, warehousing, and retail. The BMW partnership focuses on tasks like automotive body shop, sheet metal handling, and warehouse operations. Long-term, they envision humanoids assisting in homes and a wide range of other sectors.

Agility Robotics

Key Information
  • Founded: 2015 (spun out of Oregon State University).
  • Headquarters: Corvallis, Oregon, USA.
  • Key Figures: Damion Shelton (Co-founder & CEO), Jonathan Hurst (Co-founder & Chief Robot Officer).
  • Flagship Humanoid: Digit. (Earlier robot: Cassie).
  • Valuation: Potentially $1.75 billion pre-investment ahead of an anticipated $400M funding round in April 2025.
  • Official Website: agilityrobotics.com
Origin & Founding Vision

Agility Robotics was founded in 2015, spinning out of research at Oregon State University's Dynamic Robotics Laboratory. Their vision is to build robots that can work in human spaces, focusing on practical applications in logistics and warehouse automation.

Philosophy & Approach

Agility Robotics focuses on bipedal locomotion for navigating human-centric environments. Their design philosophy prioritizes practicality and reliability for specific industrial tasks, particularly in logistics and material handling. They aim for robots that can work alongside humans and integrate into existing workflows, emphasizing cooperative safety.

Leadership

Co-founded by Damion Shelton (CEO) and Jonathan Hurst (Chief Robot Officer and Professor at Oregon State University), who brings deep expertise in legged locomotion. Melonee Wise is the Chief Product Officer.

Key Robots & Technologies

Digit is their primary humanoid robot, designed for logistics tasks like moving totes and packages in warehouses. It features a bipedal design (approx. 1.7m tall, 42kg) for navigating stairs and uneven terrain, and arms for carrying objects up to 16 kg (35 lbs). Their earlier robot, Cassie, was known for its dynamic walking and running capabilities.

Digit
  • Type: Bipedal humanoid robot (approx. 1.7m tall, 42kg).
  • Capabilities: Designed for material handling, package delivery, and other logistics tasks in warehouses and fulfillment centers (e.g., tote consolidation). Can walk, navigate obstacles, climb stairs, and carry packages up to 16 kg (35 lbs). Features advanced grasping capabilities and a sensor suite (LiDAR, depth cameras) for navigation and object recognition.
  • Design: Human-like form to operate in spaces designed for humans. Features digitigrade legs for efficient locomotion. Linux-based OS, dual Intel i7 processors.
  • Autonomy & AI: Operates independently using AI and LLMs for environmental learning and task adaptation. Can autonomously dispatch AMRs.
  • Enhancements (2025): Extended battery life (up to 4 hours), autonomous docking to charging stations, additional safety features (Category 1 stop, Safety PLC, E-stops), redesigned limbs and end effectors.
  • Deployment: Nearly 100 units deployed with clients like Amazon, Spanx (via GXO Logistics).
Cassie
  • Type: Bipedal robot (legs only).
  • Legacy: Known for its agility and setting a Guinness World Record for the fastest 100-meter dash by a bipedal robot. Served as a research platform that informed the development of Digit.
Long-Term Vision / Role of Humanoids

Agility Robotics aims to build robots that augment the human workforce, taking on physically demanding or repetitive tasks in logistics and supply chain operations. Their focus is on creating practical, reliable robots that can seamlessly integrate into existing infrastructure designed for humans, complementing existing automation.

Funding & Investors

Agility Robotics has raised significant funding, including a $150 million Series B round in 2022 with investors like DCVC, Playground Global, and Amazon's Industrial Innovation Fund. In April 2025, reports indicated the company was set to secure an additional $400 million, led by WP Global Partners with participation from SoftBank Group Corp, which would elevate its pre-investment valuation to $1.75 billion.

Recent Developments (2024-2025)

Continued deployment and testing of Digit in partner facilities, notably with Amazon and GXO Logistics (for Spanx) for warehouse tasks like tote handling. Announced significant enhancements to Digit in early 2025, including expanded battery life, autonomous docking, and improved safety features. Focus on scaling production of Digit and enhancing its software for more autonomous and versatile operation, including better integration with AMRs via their Arc cloud robotic platform.

Target Audience & Use Cases

Primarily targets the logistics, warehousing, and supply chain industries. Use cases include package handling, tote moving (e.g., from AMRs to conveyors), and other material handling tasks within fulfillment centers and distribution hubs.

Sanctuary AI

Key Information
  • Founded: 2018.
  • Headquarters: Vancouver, Canada.
  • Key Figures: Geordie Rose (Co-founder & CEO), Suzanne Gildert (Co-founder & CTO).
  • Flagship Humanoid: Phoenix (currently 7th generation bipedal, with an 8th generation featuring wheeled mobility).
  • AI Platform: Carbon™.
  • Official Website: sanctuary.ai
Origin & Founding Vision

Founded in 2018 by Geordie Rose and Suzanne Gildert, Sanctuary AI is on a mission to create the world's first human-like intelligence in general-purpose robots. Their vision is to help people work more safely, efficiently, and sustainably by addressing labor challenges.

Philosophy & Approach

Sanctuary AI's approach focuses on developing "human-like intelligence" through their AI control system, Carbon™, which powers their Phoenix humanoids. They emphasize highly dexterous hands and fine manipulation capabilities, believing that mastering physical tasks is key to general-purpose robotics. Their philosophy involves integrating best-in-class hardware and AI, and they see tactile sensing as critical. The 8th gen Phoenix adopted wheeled mobility for stability and efficiency, focusing on upper-body manipulation.

Core Tenets
  • Human-like Intelligence: Aiming to replicate human cognitive abilities in their Carbon™ AI platform to enable natural human-robot interaction.
  • Dexterous Manipulation: Strong focus on hand dexterity (30+ joints for fluid movement), fine manipulation, and touch sensing as crucial for performing real-world work.
  • General-Purpose Robots: Developing robots capable of performing a wide range of tasks across various industries.
  • Embodied AI: Believes the best way to build AGI is through software controlling sophisticated robots with human-like senses and actions.
  • Iterative Design: Phoenix has undergone multiple generations of development, with the 8th gen moving to wheeled mobility for certain advantages while maintaining focus on upper-body tasks.
Leadership

Led by co-founders Geordie Rose (CEO), known for his previous work with D-Wave Systems, and Suzanne Gildert (CTO), a prominent researcher in AI and robotics.

Key Robots & Technologies

Phoenix is their flagship general-purpose humanoid robot, now in its 7th (bipedal) and 8th (wheeled) generations. It features highly dexterous, human-like hands with up to 21 degrees of freedom per hand and advanced tactile sensing. The robots are controlled by Carbon™, Sanctuary AI's cognitive architecture and software platform. Phoenix is approx. 1.65m tall, weighs around 55kg, and has a battery life of about 4 hours.

Phoenix Humanoid Robot
  • Generations: 7th generation featured bipedal locomotion; 8th generation utilizes a wheeled base for enhanced stability and energy efficiency, focusing on upper-body manipulation.
  • Height & Weight: Approximately 1.65 m; Around 55 kg.
  • Degrees of Freedom: 30+ joints for fluid movement. Hands have up to 21 DoF per hand.
  • Battery Life: Approximately 4 hours per charge.
  • Hands: Industry-leading dexterous hands with advanced tactile sensing, crucial for fine manipulation tasks.
  • AI Control: Powered by Carbon™, an AI control system designed for human-like intelligence and learning. Enables understanding and response to human cues.
  • Sensors: Integrates vision systems and touch sensors for tactile feedback.
Carbon™ AI Platform
  • The cognitive architecture and software platform that enables Phoenix to learn and perform tasks. Designed to allow robots to understand and interact with their environment in a human-like way.
Long-Term Vision / Role of Humanoids

Sanctuary AI's explicit goal is to create "human-like intelligence" in general-purpose robots, essentially aiming for Artificial General Intelligence (AGI) embodied in a humanoid form. They believe this will allow robots to perform physical labor across virtually every industry and address global labor shortages.

Funding & Investors

Sanctuary AI has raised significant funding from investors including Accenture Ventures, Bell Canada, Magna, Microsoft, and Verizon. They have formed partnerships with companies like Canadian Tire for commercial deployment.

Recent Developments (2024-2025)

Successfully completed the first commercial deployment of a general-purpose humanoid robot system at a Mark's (part of Canadian Tire Corporation) retail store. Acquired IP assets related to touch and grasping technologies to enhance Phoenix's manipulation capabilities. Introduced the 8th generation Phoenix with wheeled mobility. Focused on advancing the Carbon™ AI platform for improved learning, task execution, and seamless human-robot interaction.

Target Audience & Use Cases

Targeting a wide variety of industries facing labor shortages, including service and collaborative work environments. Initial commercial deployments include retail environments for tasks like stocking shelves, cleaning, and customer assistance. Future applications could span logistics, manufacturing, healthcare, and more.

Apptronik

Key Information
  • Founded: 2016 (spun out of the Human Centered Robotics Lab at the University of Texas at Austin).
  • Headquarters: Austin, Texas, USA.
  • Key Figures: Jeff Cardenas (Co-founder & CEO), Dr. Nick Paine (Co-founder & CTO).
  • Flagship Humanoid: Apollo.
  • Official Website: apptronik.com
Origin & Founding Vision

Apptronik spun out of the University of Texas at Austin in 2016, building on extensive experience in developing robotic systems, including work with NASA on robots like Valkyrie. Their vision is to build robots that can work alongside humans and augment the human workforce in various industries.

Key Details
  • NASA Roots: The founders previously worked on NASA's Valkyrie humanoid robot, participating in the DARPA Robotics Challenge. This experience shaped their design philosophy for robust and capable humanoids.
  • Iterative Development: Apptronik has developed numerous robotic systems, iterating towards their general-purpose humanoid, Apollo. Their philosophy has been "show versus tell," focusing on building and demonstrating capabilities.
Philosophy & Approach

Apptronik's philosophy is "it is not human against the machine, but human + machine," emphasizing the collaborative role of robots. They focus on a human-centered design, aiming for robots that are safe, user-friendly, and can seamlessly integrate into human environments and workflows. They believe in building the fundamental building blocks of robotics to create versatile platforms. Apollo is designed for friendly interaction, mass manufacturability, high payloads, and safety.

Core Tenets
  • Human-Robot Collaboration: Robots are seen as tools to expand and complement human skills rather than replace workers, designed to work alongside humans in dynamic environments.
  • Human-Centered Design: Apollo is designed with ergonomics and user-friendliness in mind for intuitive interaction.
  • Versatility and Adaptability: Aiming for robots that can perform a wide range of tasks across multiple industries by being able to do many different things.
  • Real-World Application: Focused on solving practical problems in areas like logistics, manufacturing, and eventually beyond. Initial testing involves repetitive tasks in intralogistics.
Leadership

Led by Co-founder and CEO Jeff Cardenas, and Co-founder and CTO Dr. Nick Paine.

Key Robots & Technologies

Apollo is their flagship humanoid robot, unveiled in August 2023. It is designed for industrial work in warehouses and manufacturing plants, featuring a modular design and an emphasis on safety and human-robot interaction. Apollo utilizes linear actuators and stereo vision, stands approximately 1.7m tall, weighs around 60kg, has 28+ joints, and an estimated 5-hour battery life.

Apollo
  • Height & Weight: Approx. 1.7m (5'8"), 60-73 kg (132-160 lbs). Original doc: 5'8" (1.73m), 160 lbs (73 kg).
  • Payload: 55 lbs (25 kg).
  • Degrees of Freedom: 28+ joints for refined motion control.
  • Battery Life: Estimated at 5 hours per charge.
  • Design: Humanoid form for working in human spaces. Modular design for adaptability. Uses linear actuators.
  • Perception: Equipped with stereo vision for 3D environmental understanding.
  • Safety: Designed for safe collaboration with human workers.
  • AI Integration: Partnering with Google DeepMind to integrate advanced AI capabilities. Mercedes-Benz employees transfer knowledge to Apollo using teleoperation and augmented reality.
Long-Term Vision / Role of Humanoids

Apptronik envisions a future where general-purpose robots are as ubiquitous as cars, helping people with work that needs doing across countless sectors. Their goal is to develop machines that enable people to reach their full potential by blurring the lines between art and technology. They aim for robots that can integrate seamlessly into environments already built around humans.

Funding & Investors

Apptronik raised an additional $53 million in its Series A round in March 2025, bringing the total Series A funding to $403 million. New investors include Mercedes-Benz, Japan Post Capital, ARK Invest, Helium-3, Magnetar, RyderVentures, and a syndicate led by Korea Investment Partners. Previous investors include B Capital, Capital Factory, and Google. Mercedes-Benz also committed a low double-digit-million-euro investment into Apptronik.

Recent Developments (2024-2025)

Announced a partnership with Mercedes-Benz in March 2024/2025 to deploy Apollo robots in their manufacturing facilities in Berlin and Hungary for tasks like transporting components and quality checks. Robots are collecting data and being trained via teleoperation. Apptronik is also testing robots with GXO Logistics and has a pilot partnership with Jabil. Collaboration with Google DeepMind continues to advance AI for their robots. Widespread use across Mercedes factories is expected by 2030.

Target Audience & Use Cases

Initial focus on logistics and manufacturing sectors (e.g., automotive with Mercedes-Benz for intralogistics tasks). Also targeting electronics manufacturing and consumer packaged goods. Future applications are envisioned in construction, oil and gas, retail, home delivery, healthcare, hospitality, and elder care. Apollo's price is expected to be between $9,000 and $100,000.

Unitree Robotics

Key Information
  • Founded: 2016.
  • Headquarters: Hangzhou, China.
  • Key Figures: Wang Xingxing (Founder & CEO).
  • Flagship Humanoids: H1, G1 (a smaller, more affordable version).
  • Other Products: Known for its quadruped robots (e.g., Go series, B series).
  • Official Website: unitree.com
Origin & Founding Vision

Unitree Robotics was founded by Wang Xingxing with a vision to make high-performance robotics, including humanoids, more accessible and affordable. They initially gained prominence with their quadruped robots before expanding into humanoids. Their philosophy emphasizes "motion before cognition" in some respects, prioritizing dynamic movement capabilities.

Philosophy & Approach

Unitree focuses on developing core robotic components (motors, reducers, controllers) in-house to reduce costs and enhance performance. They aim for extreme cost control and vertical integration to make advanced robots accessible for research, education, and eventually consumer markets. Their approach involves rapid iteration and sometimes unconventional design choices to achieve breakthroughs in cost and capability, with over 150 global patents.

Core Tenets
  • Affordability and Accessibility: A primary goal is to significantly lower the price point of advanced robots like humanoids (G1 starting around $16,000, H1 around $90,000).
  • In-House Component Development: Self-research and development of key components like motors and sensors to control costs and performance.
  • Vertical Integration: Building its own factories to lower supply chain costs and optimize product structure.
  • Motion-First Philosophy (for some aspects): Prioritizing robust and dynamic movement capabilities, sometimes described as "motion before cognition," to quickly enable physical interaction. Demonstrated G1 performing side-flips and kick-ups.
  • Openness for Development: Provides open-source software and datasets for development with some of its robots (e.g., LAFAN1 motion capture for G1), encouraging research and community engagement.
Leadership

Led by Founder and CEO Wang Xingxing.

Key Robots & Technologies

The H1 (approx. 1.8m tall, 47kg) is their full-sized general-purpose humanoid, capable of backflips and a max speed of 4.3m/s. The G1 (approx. 1.27-1.32m tall, 35-47kg), unveiled in 2024, is a smaller, more affordable humanoid robot, priced around $16,000. Both feature high joint flexibility (G1: 23-43 DoF) and are capable of complex movements like dancing and side-flips. The G1 uses UnifoLM (Unitree Robot Unified Large Model) and can have 3D LIDAR and depth cameras.

Unitree G1
  • Price: Starts around $16,000 (base configuration).
  • Height & Weight: Approx. 1.27m-1.32m tall, weighs around 35-47 kg.
  • Degrees of Freedom (DoF): Base model 23 DoF, with some versions up to 43 DoF.
  • Capabilities: Designed for flexibility, dexterous manipulation (optional hands), imitation and reinforcement learning. Can perform dynamic movements, including jumps, side-flips, kick-ups, and martial arts-like poses. Battery life around 2 hours (9000 mAh). Can move up to 7.2 km/h.
  • Technology: Uses Unitree's UnifoLM (Unitree Robot Unified Large Model) for intelligence. Features hollow electrical routing. Optional 3D LIDAR LIVOX-MID 360 sensor and 360° detection perception depth cameras.
  • Target Market: Robotics research and development, education, potentially light industrial or entertainment. Not designed as a direct human labor replacement due to its smaller stature.
Unitree H1
  • Price: Base price around $90,000. (Original doc, some sources say $16k for H1, which is likely G1) Height approx 1.8m, weight 47kg.
  • Capabilities: A larger, more powerful humanoid aimed at general-purpose applications. Showcased performing tasks like folk dancing and backflips. Max movement speed of 4.3m/s. Holds a Guinness World Record for sprinting speed.
Long-Term Vision / Role of Humanoids

Unitree aims to drive robots from the lab to the broader market by drastically reducing costs. While acknowledging the current limitations of AI for complex reasoning, their focus is on creating physically capable and affordable platforms that can be enhanced with improving AI over time. They envision robots becoming more integrated into daily life and aim to bring a breakthrough in humanoid capabilities by 2026.

Funding & Investors

Unitree Robotics has received funding from investors, though specific rounds and valuations are not always as widely publicized as some US counterparts. Their strategy of in-house development and cost control is a key part of their financial approach. They have over 150 global patents.

Recent Developments (2024-2025)

Unveiled the G1 humanoid robot in 2024 with a disruptive price point. Showcased G1 and H1 robots performing human-like dance movements and impressive agility like side-flips (G1) in early 2025. An "agile upgrade" for G1 in early 2025 improved mobility, navigation on rough terrain, and AI algorithms. Unitree announced a robot boxing fight for May 2025, likely featuring G1 or H1. Continued emphasis on mass production readiness and making advanced robotics more accessible, showcasing at CES 2025.

Target Audience & Use Cases

The G1 targets research institutions, universities, and developers due to its affordability and open development potential (EDU version). Also suitable for entertainment, business promotions, search-and-rescue, and potentially light industrial automation or hazardous condition applications. The H1 is aimed at more general-purpose applications.

1X Technologies

Key Information
  • Founded: 2014 (as Halodi Robotics).
  • Headquarters: Moss, Norway, with presence in San Francisco Bay Area, USA.
  • Key Figures: Bernt Øyvind Børnich (Founder & CEO).
  • Flagship Humanoids: NEO (bipedal, for home/consumer use), EVE (wheeled, for industrial/commercial use). Latest iteration of NEO is NEO Gamma.
  • Official Website: 1x.tech
Origin & Founding Vision

Founded in 2014 as Halodi Robotics by Bernt Øyvind Børnich, 1X Technologies aims to create an "abundant supply of labor via safe, intelligent robots." Their vision is for humanoid robots to integrate into everyday life, particularly in homes, to assist with daily tasks and coexist with humans.

Philosophy & Approach

1X's philosophy centers on developing humanoids "alongside humans, not in isolation," believing that real-world home environments provide crucial data for AI development. Safety, practicality, and affordability are key tenets. They employ an embodied AI approach, integrating AI into physical bodies, and focus on creating robots that can learn and adapt in human spaces. Their NEO Gamma features a soft, 3D-printed fabric skin to feel more approachable in homes.

Core Tenets
  • Home-Centric Development: Believes humanoids need to live and learn among humans in homes to develop true intelligence and understand real-world nuances.
  • Safety and Coexistence: Prioritizing the safe operation of robots in human environments.
  • Embodied AI: Integrating AI software with physical robotic bodies, often co-training algorithms with partners like OpenAI.
  • Practical Assistance: Focusing on robots that can perform useful household chores and provide assistance.
  • Approachable Design: NEO Gamma's fabric skin is an example of designing robots to better blend into domestic settings and be psychologically more acceptable.
Leadership

Led by Founder and CEO Bernt Øyvind Børnich. Christoph Kohstall, founder of Kind Humanoid (acquired by 1X), also joined the team.

Key Robots & Technologies

NEO is their bipedal humanoid designed for home assistance, with NEO Gamma being the latest iteration featuring a softer design and improved AI. EVE is their wheeled humanoid robot platform, initially deployed for industrial tasks like security and logistics. They focus on developing their own AI software, including for vision, language, and motion, and utilize reinforcement learning, with collaboration from OpenAI.

NEO (including NEO Beta, NEO Gamma)
  • Type: Bipedal humanoid robot for consumer/home use.
  • Design (NEO Gamma): Features a 3D-printed nylon fabric "skin" for a softer, more approachable appearance. Lightweight design (approx. 66 lbs / 30 kg). Tendon-driven hands for gentle manipulation.
  • Capabilities: Designed for household chores (tidying, cleaning), interaction with humans, and navigating home environments. Capable of carrying up to 44 lbs (20 kg). Battery life of 2-4 hours.
  • AI: Utilizes embodied AI, learning from real-world interaction in homes. Collaboration with OpenAI for AI functions.
EVE
  • Type: Wheeled humanoid robot for commercial/industrial applications.
  • Capabilities: Deployed for tasks in security and logistics. Utilizes a self-balancing wheeled base.
Long-Term Vision / Role of Humanoids

1X envisions a future where everyone has a robot helper at home, like "Rosey the Robot" or Baymax. Their goal is to create an "abundant society" by meeting global labor demands through commercially scaled, safe, and advanced androids that can perform everyday tasks.

Funding & Investors

Raised $100 million in a Series B funding round in January 2024, led by EQT Ventures, with participation from Samsung NEXT and existing investors like OpenAI, Tiger Global, and Sandwater. Total funding exceeds $125 million.

Recent Developments (2024-2025)

Unveiled NEO Gamma in February 2025, an improved version of their home humanoid. Acquired California-based startup Kind Humanoid in January 2025 to bolster their team and expertise in home-focused humanoids. Launched a private early-access program to test NEO robots in homes by late 2025. Focused on scaling their data collection strategy for Embodied AI.

Target Audience & Use Cases

NEO is targeted at the consumer market for everyday home assistance, including household chores and companionship. EVE serves enterprise clients in logistics and guarding. Potential future applications for NEO include elderly care. Pricing for NEO is anticipated to be comparable to an "affordable car."

Engineered Arts

Key Information
  • Founded: 2004/2005.
  • Headquarters: Falmouth, Cornwall, UK. Recently restructured as a U.S. entity (December 2024) with a presence in Redwood City, California.
  • Key Figures: Will Jackson (Founder & CEO).
  • Flagship Humanoids: Ameca (advanced expressive humanoid), Mesmer (hyper-realistic human-like robots), RoboThespian (earlier entertainment robot).
  • Operating System: Tritium.
  • Official Website: engineeredarts.co.uk (or .com for US entity)
Origin & Founding Vision

Engineered Arts was founded by Will Jackson in Cornwall, UK, over 20 years ago, initially creating mechanical actors for entertainment and science centers, with RoboThespian being an early success. Their vision evolved towards creating highly realistic and expressive humanoid robots for interaction, entertainment, education, and research.

Philosophy & Approach

Engineered Arts focuses on creating human-centric robots that foster natural and intuitive interactions, with a motto "Be wow!". They emphasize lifelike facial expressions, smooth movements, and the integration of AI for engaging communication. They aim to make humanoid robots more accessible (full-sized and desktop versions) and bring them closer to everyday practical use by enhancing dexterity and locomotion.

Leadership

Led by Founder and CEO Will Jackson, a self-taught engineer who has guided the company's development in humanoid robotics for over 20 years. Matthew Bellamy (Muse frontman, Helium-3 Ventures partner) joined the board as an observer in December 2024.

Key Robots & Technologies

Ameca is their most advanced humanoid, known for its highly expressive face and human-like interactions, designed as a platform for AI development. Mesmer robots are hyper-realistic, designed to be anatomically accurate and highly expressive. RoboThespian is an earlier interactive entertainment robot. All run on their Tritium operating system.

Ameca
  • Type: Advanced expressive humanoid robot.
  • Capabilities: Designed for human-robot interaction, featuring a wide range of facial expressions, natural gestures, and AI-driven conversational abilities. Serves as a development platform for AI and machine learning.
  • Design: Deliberately non-human skin tone and features to be clearly identifiable as a robot, while being highly expressive. Modular design.
Mesmer
  • Type: Hyper-realistic humanoid robots.
  • Capabilities: Designed to be anatomically accurate with realistic skin, hair, and facial expressions, often modeled on real people or characters. Used for entertainment, museums, and attractions.
RoboThespian
  • Type: Interactive entertainment and educational robot.
  • Legacy: An earlier successful product line used in science centers, museums, and events worldwide for engaging audiences.
Tritium OS
  • A cloud-based operating system developed by Engineered Arts to control their robots, manage content, and facilitate interaction.
Long-Term Vision / Role of Humanoids

Engineered Arts envisions a world where the virtual seamlessly integrates into everyday life through humanoid robots. Their robots are designed to support, entertain, inform, and educate, providing a human-centric vision of AI-driven technology. They aim to enhance dexterity and locomotion to bring humanoids closer to everyday functionality, potentially creating Digital Employees and Non-Human Workers for industries like healthcare and customer service.

Funding & Investors

Secured $10 million in a Series A funding round in December 2024, led by Helium-3 Ventures, alongside restructuring as a U.S. entity. This brought their total funding to $16.2 million. Other investors include AppDirect Chairman and CEO Nicolas Desmarais, Belvoir Investments, ThirtySeven Holdings Inc., and Figueira Capital.

Recent Developments (2024-2025)

Restructured as a U.S. entity and secured $10 million in Series A funding (December 2024) to scale production of full-sized and desktop robots, refine products, expand cloud-based AI services, and enhance dexterity and locomotion. Plans to launch a virtual robot character platform and expand their Redwood City, CA location.

Target Audience & Use Cases

Targets entertainment, education, research, and business applications. Robots are used in science centers, theme parks, corporate events (e.g., by GlaxoSmithKline), museums, and as platforms for AI research and human-robot interaction studies. Examples include exhibits at The Sphere in Las Vegas and the Computer History Museum in Mountain View, California.

Fourier Intelligence

Key Information
  • Founded: 2015.
  • Headquarters: Shanghai, China, with global operations.
  • Key Figures: Alex Gu (Co-founder & Group CEO).
  • Flagship Humanoids: GR-1 (launched 2023), GR-2.
  • Primary Focus (Historically): Rehabilitation robotics, expanding into general-purpose humanoids.
  • Official Website: fourierintelligence.com
Origin & Founding Vision

Fourier Intelligence was founded in 2015, initially focusing on developing intelligent robotics for medical rehabilitation and exoskeletons, providing solutions to over 2,000 hospitals and institutions in over 40 countries. They have since leveraged this expertise to expand into the development of general-purpose humanoid robots.

Philosophy & Approach

Fourier Intelligence's philosophy is rooted in applying their deep understanding of human biomechanics, motion control, and sensor technology from the rehabilitation sector to create capable and versatile humanoid robots. They aim to mass-produce humanoids for a variety of applications.

Leadership

Led by Co-founder and Group CEO Alex Gu.

Key Robots & Technologies

The GR-1 was their first mass-produced general-purpose humanoid, unveiled in 2023 with capabilities for dynamic walking, obstacle avoidance, and carrying loads (up to 50kg). The GR-2 is a more recent iteration, showcased with enhanced features. These robots benefit from Fourier's experience in actuators and control systems developed for rehabilitation.

GR-1 / GR-2 Humanoid Robots
  • Type: General-purpose humanoid robots.
  • Capabilities: Designed for tasks such as assisting the elderly, providing companionship, and potentially light industrial or service roles. Capable of stable walking, carrying objects (e.g., GR-1 can carry loads up to 50kg), and navigating environments.
  • Technology: Leverages Fourier's expertise in high-performance actuators, sensors, and motion control algorithms derived from their rehabilitation robotics work.
Long-Term Vision / Role of Humanoids

Fourier Intelligence aims to be a leading global provider of humanoid robots, with a goal of mass production. They envision humanoids playing a significant role in various sectors, including healthcare, education, and home assistance, augmenting human capabilities.

Funding & Investors

Fourier Intelligence has received substantial investment from major firms, including SoftBank Group and Saudi Aramco's Prosperity7 Ventures, to support R&D and mass production efforts for their humanoid robots.

Recent Developments (2024-2025)

Ongoing demonstrations of their GR series humanoids performing various tasks. Focus on scaling up manufacturing capabilities for mass production. Expanding partnerships for deploying humanoids in different application scenarios. The GR-2 was noted as a key robot to watch in 2025.

Target Audience & Use Cases

Initially targeting healthcare (building on their rehabilitation roots), elder care, research, and education. Long-term plans include broader service industries, light manufacturing, and potentially home companionship.

UBTECH Robotics

Key Information
  • Founded: 2012.
  • Headquarters: Shenzhen, China.
  • Key Figures: Zhou Jian (Founder, Chairman, CEO).
  • Flagship Humanoids: Walker S (industrial), Tien Kung Xingzhe (research, approx. $41,200).
  • AI Platform: BrainNet (for humanoid collaboration), Internet of Humanoids (IoH).
  • Official Website: ubtrobot.com
Origin & Founding Vision

UBTECH Robotics, founded in 2012, is one of the pioneering humanoid robot companies from China. They aim to bring intelligent robots into every home and business, focusing on R&D, manufacturing, and sales of AI-powered service robots and solutions. Their vision includes creating a future where humanoid robots are integral to daily life and industrial automation.

Philosophy & Approach

UBTECH emphasizes "AI at its core and humanoid robots as the carrier." They focus on full-stack R&D from core components to software and algorithms, including swarm intelligence for multi-robot collaboration. They aim for mass production and practical deployment in industrial settings, moving from "usable" to "highly efficient" while reducing costs. Significant investment in R&D, holding over 2,680 global patents by end of 2024.

Core Tenets
  • AI-Centric Development: Leveraging AI, including large models like DeepSeek-R1 for high-level planning ("big brain") and Transformer models for real-time motion control ("small brain").
  • Full-Stack Innovation: In-house development of core technologies including servo drives, motion control algorithms, and robot operating systems.
  • Swarm Intelligence: Developing systems like BrainNet and IoH for multiple humanoids to collaborate on complex tasks.
  • Industrial Application Focus: Prioritizing deployment in manufacturing and logistics, with robots like Walker S designed for factory environments.
  • Cost Reduction & Mass Production: Aiming for industrial-scale production within 1-2 years (as of early 2025) and continuous cost optimization.
Leadership

Led by Founder, Chairman, and CEO Zhou Jian.

Key Robots & Technologies

Walker S is an industrial humanoid (1.7m tall, 65kg) designed for manufacturing tasks, featuring force-compliant drive joints and capable of operating on fast-moving production lines. Walker S Lite is a more compact version. Tien Kung Xingzhe (1.7m tall) is an upgraded research model ($41,200) with enhanced strength, stability, and ability to integrate features like lidar and Nvidia Jetson Orin. They also developed the Tiangong Ultra humanoid.

Walker S / S Lite
  • Type: Industrial-grade humanoid robot. Walker S is 1.7m tall, 65kg.
  • Capabilities: Automates complex tasks in manufacturing (sorting, transport, assembly). Features force-compliant drive joints and rigid-flexible coupling hybrid structures for stability on production lines. Battery autonomy approx. 2.5 hours, 41 DoF. Deployed in fleets for collaborative tasks.
  • Walker S Lite: A more compact and cost-effective version.
Tien Kung Xingzhe
  • Price: Approx. US$41,200.
  • Height: 1.7m.
  • Capabilities: Upgraded version of Tien Kung model, designed for research with enhanced strength and stability. Can integrate additional lidars, Nvidia Jetson Orin, and five-fingered hands. Walking speed up to 10 km/h, can operate on varied terrains.
Tiangong Ultra
  • Developed by the Beijing Humanoid Robot Innovation Center (initiated by UBTECH), won the world's first half-marathon for humanoid robots in April 2025.
Long-Term Vision / Role of Humanoids

UBTECH envisions humanoid robots as the "ultimate realization of embodied intelligence," transforming industries and daily life. They aim to achieve industrial-scale production and see China's humanoid market growing significantly by 2030. The next-gen Walker S2 aims for self-charging, enhanced dexterity, and lower costs.

Funding & Investors

UBTECH is a publicly listed company. They invest nearly half of their revenue into R&D. Strategic partnerships with companies like Foxconn and automotive manufacturers like Geely (Zeekr) and BYD for deployment and development.

Recent Developments (2024-2025)

Deployed Walker S1 robots at Geely's Zeekr 5G smart factory in March 2025, demonstrating multi-robot collaborative training and swarm intelligence via their BrainNet system. Mass deployment of Walker S1 in BYD's EV plants slated for Q2 2025. UBTECH announced plans to mass-produce humanoids by 2025, aiming to deliver 500-1,000 units in the year. Launched the Tien Kung Xingzhe for pre-order in March 2025, with deliveries expected Q2 2025.

Target Audience & Use Cases

Primary focus on industrial applications, particularly automotive manufacturing (assembly, logistics tasks at Zeekr, BYD, Audi-FAW). The Tien Kung Xingzhe targets research applications. Future applications include broader industrial automation, services, and potentially home use.

Xiaomi

Key Information
  • Project Announced: CyberOne unveiled August 2022.
  • Headquarters: Beijing, China (Xiaomi Corporation).
  • Key Figures: Lei Jun (CEO of Xiaomi).
  • Flagship Humanoid: CyberOne.
  • Parent Company: Xiaomi Corporation.
  • Official Website: mi.com/global/discover/robotics
Origin & Founding Vision

Xiaomi unveiled its first humanoid robot, CyberOne, in August 2022, signaling its ambition to expand beyond consumer electronics into intelligent robotics. The vision is to explore the future of human-robot interaction and advanced robotics.

Philosophy & Approach

Xiaomi's robotics development focuses on AI-human interaction, sensorimotor coordination, and creating robots capable of perceiving 3D space and recognizing human emotions and gestures. Their approach leverages their expertise in AI and hardware manufacturing from their consumer electronics business. While still in early stages, CyberOne demonstrates their commitment to this field.

Key Robots & Technologies

CyberOne is a bipedal humanoid robot standing 177cm tall and weighing 52kg. It features a Mi-Sense depth vision module for 3D space perception, AI interaction algorithms for recognizing individuals and gestures, and MiAI emotion recognition capabilities (45 human emotions). It has 21 degrees of freedom and high torque joints (up to 300Nm).

CyberOne
  • Height & Weight: 177cm, 52kg.
  • Mobility: Bipedal with 21 degrees of freedom in motion, 0.5ms response speed per DoF for smooth movement.
  • Perception: Mi-Sense depth vision module for 3D space perception, AI interaction algorithm for recognizing individuals, gestures, and expressions.
  • Emotional Intelligence: MiAI emotion recognition capabilities to identify 45 human emotions and 85 types of environmental sounds.
  • Mechanics: Features high torque joints (up to 300Nm) for stability. OLED face display.
Recent Developments (2024-2025)

Xiaomi continues to develop CyberOne, with ongoing advancements in mobility, AI-driven perception, and new capabilities. However, in March 2025, Xiaomi denied rumors of mass production for CyberOne, stating the reports were inaccurate. The focus remains on R&D and future enhancements, including potential integration with IoT devices.

Target Audience & Use Cases

Currently, CyberOne serves primarily as a research and development platform, showcasing Xiaomi's technological capabilities. Future prospects could include applications in smart homes, assistance, and potentially light manufacturing if development progresses towards mass production and enhanced capabilities.