Humanoid Robotics Database

Comprehensive field database covering companies, market economics, technology, dexterous hands, and investment landscape

0
Companies
0
Robot Models
22
Countries
$38B
TAM 2035
Company ↕Country ↕Robot Model(s)SegmentWebsite

Top 10 Most Notable Vendors – Q3 2025

Scored 1-10 across technology advantage, commercial momentum, AI capabilities, hardware IP, manufacturing, funding, talent, patents, gov support, supply chain, and cost competitiveness.

CompanyCountrySafetyTechCommercialAIHW IPMfgFundingTalentPatentsGovSupplyCostTotal

Market Size & Growth

$1.55B
2024 Global Revenue (GVR)
~9,000
2024 Installed Fleet (est.)
~1%
Share of Industrial Robot Mkt
$4B
2030 Forecast (GVR Base)
$38B
2035 Forecast (Goldman Bull)
$5T
2050 TAM (Morgan Stanley)
>1B
2050 Robot Population (MS)
17-35%
CAGR Range 2024-2035

Cost Curve Collapse

2018 — Atlas Research Unit

≈ US $1,000,000

Research Only

2023 — Optimus Prototype (Tesla est.)

≈ US $150,000

-85% vs 2018

2024 — Unitree G1 Street Price

≈ US $5,600

-99.4% vs 2018

2025 — Unitree R1

≈ US $5,900

Consumer-accessible

2032 — Projected Mass Volume

≈ US $10,000

Target

Enterprise ROI Snapshot

MetricHuman Picker (Night Shift)Humanoid (RaaS)
Annual CostUS $49,000US $21,600
Available Hours/Year1,9007,000
Cost per Productive HourUS $26US $3
Break-even< 9 months at premium lease rates

Capital Flows

Figure AI — Series C (2025)

Raised: US $1.5B | Post-money: ~$39.5B

Nvidia, SoftBank, Tiger Global

1X Technologies — Series A2 (2024)

Raised: US $1B | Post-money: ~$5B

OpenAI, NVIDIA, EQT Ventures, Tiger Global

Agility Robotics — Series C (2022)

Raised: US $150M | Private

DCVC, Amazon, Sony Innovation Fund

Consumer Willingness-to-Pay

Wait until ~$20K (2030)
43%
Buy at ~$50K within 2 yrs
35%
Rent at ~$1,500/mo
17%
Not interested
3%
See no use
2%

Purchase Timeline

Within 3 months
~295
Immediately ("Now")
~210
Within 6 months
~235
Within 12 months
~115
Within 18 months
~100
Within 24 months
~70

Top Home Tasks (What People Want First)

Clean the house
≈33%
Do the laundry
≈28%
Work in the garden
≈24%
Provide companionship
≈22%
Organise & declutter
≈19%
Assist with maintenance
≈17%
Chef, security, dishes…
≈15%

Consumer Segments

Tech-Enthusiast Pioneers (~15%)

Latest hardware, social-media bragging rights. Willing to pay $50K+.

Limited-edition runs, premium support

Early Economic Adopters (~35%)

Tangible labour savings (cleaning, errands). Budget ≤$50K.

ROI messaging, finance plans

Value-Conscious Pragmatists (~40%)

Car-level pricing (~$20K), reliability over novelty.

Scale economics, durability, service network

Sceptics (<5%)

Need convincing use-cases and proven track record.

Target later with success stories

Feasibility Matrix

TaskTech Readiness 2025Blocking GapsLikely Timeline
Floor vacuum/mopMediumBiped balance; battery lifePilots 2026-27
Laundry handlingLowReliable cloth grasp & fold2028-30
Basic cookingLowHeat-safe manipulators; food-safe2029-32
Garden weeding/mowingLowWeather sealing; outdoor localisation2027-29
Child/elder companionshipHigh (AI speech)Safety certification; trust frameworksIncremental via OTA

Enterprise Willingness-to-Pay

Buy at ~$50K (1-2 yrs)
≈64% (~540)
Wait until ~$20K (3-5 yrs)
≈41% (~350)
Don't see use
≈19% (~160)

Top Industrial Tasks

Cleaning
≈60%
Moving totes & boxes
≈50%
Managing inventory
≈35%
Warehouse errands
≈27%
Office support
≈21%
Security
≈15%
Making & serving lunch
≈15%

Sector Readiness Matrix

IndustryNear-term Use CasesPain-point FitBarrier / Enabler
E-commerce & 3PLTrailer unloading, aisle cleaningHighROI obvious; unions watch safety
ManufacturingLine-side material deliveryHighNeeds CE/UL safety sign-off
Facilities MgmtNight-shift cleaning, securityMediumLift-integration & BMS APIs
HealthcarePharmacy runs, waste removalMediumStrict hygiene, patient trust
Hospitality & RetailRoom service, shelf restockMediumImage-sensitive; robot must be quiet
Oil & GasInspection, valve turningHighEX certification; harsh environment
ConstructionMaterial ferry, site scanningLowOutdoor durability, rough terrain

Early Commercial Proof Points

GXO × Agility (Digit)

Tote handling & trailer unloading. Multi-year paid contract, KPI ≥ 800 cases/hr. Live Q4-2024.

BYD × UBTECH (Walker S1)

Assembly-line material shuttle. 500-unit pre-order; 16h/day uptime target.

Austin Retail Trial (Unitree G1)

In-store marketing mascot. US $16K unit cost; remote-piloted to satisfy insurance.

BMW Auto OEM (Figure 01)

End-of-line inspection. Proof-of-concept; part of funding rationale.

Robot Hands — The Critical Bottleneck

27
Human Hand Bones
30+
Human Hand Muscles
17,000
Tactile Receptors/Hand
21 DoF
Human Hand Degrees of Freedom
40 kg
Human Max Grip Force
10-15 DoF
Typical Robot Hand Today

Hand Actuation Approaches

ApproachExamplesProsCons
Actuators in Palm (Direct Drive)Figure AI, Unitree (Dex5-1)Simple mechanics, high precision, local sensor easeHeavier hand, heat issues, bulkier design
Wire-Driven (Tendon/Cable)Tesla Optimus, 1X (NEO)Lightweight, backdriveable, biomimetic efficiencyBacklash/friction, maintenance, control complexity
Pneumatic (McKibben muscles)Clone RoboticsBetter embodiment of nature, compliantCompliance issues, hard to control precisely

Actuation Comparisons (Body Joints)

Hydraulic

High design freedom, brakes/clutches/springs as fluid circuit. But: leaks unacceptable indoors, hard to miniaturize. "Persian rug problem."

Electric (QDD)

Quasi-direct drives: high-torque PM motor + low-ratio transmission + sensing. Default topology for humanoids. Backdriveable, efficient.

Tendon (Large Joints)

Fine at slow speeds. At faster walking/running: heat up, fatigue, fail. Trend is direct actuation with lowest gear ratio possible.

Cycloidals vs Harmonics

Cycloidals: backdriveable but hard to make light, tough to miniaturize. Harmonics: dominant for hips/knees/shoulders, compact, low backlash.

Key Performance Factors for Future Hands

Reliability

Endurance under cyclic loads is non-negotiable. Wire-driven: cable fatigue. Palm actuators: overheating.

Dexterity

High DoF and fluid motion. Wire-driven approaches often lead due to lighter construction.

Strength

Balanced grip force (8-40 kg) without excess power draw.

Backdriveability

Allows compliance for safety. Inherent in wire systems. Prevents damage from external forces.

Tactile Sensing

SensoBright: 400+ data points per sensor, light-based, -40°C to +90°C, waterproof.

Technical Bottlenecks (2025 Status)

BottleneckStatus 2025Risk if UnresolvedMitigation
Manipulation reliability≈70% simple graspsTask failures kill ROIMulti-modal AI + force sensors
Battery energy density3h shift limitFrequent charges → downtimeFast-swap packs, solid-state cells
Fail-safe perceptionGood lighting onlyAccidents, insurance push-backLiDAR + RGB-D fusion + ISO cert

Dexterous Hands — Comprehensive Database

27
Human Hand Bones
21 DoF
Human Hand Freedom
17,000
Tactile Receptors/Hand
40 kg
Human Max Grip Force
12+
Commercial Hands
7+
Open-Source Designs
7+
Tactile Sensor Platforms
$314–$100K+
Price Range

Commercial / Closed-Source Dexterous Hands

Hand ↕CompanyCountryDoFWeightForceActuationTactileSelf-LockPriceStatus

Commercial Hand Detail Cards

SharpaWave — $50,000/hand

Sharpa Robotics (Singapore/Shanghai) — CES 2026 Innovation Award

22 DoF via 22 proprietary multi-mode power modules. 1:1 human scale using golden ratio (0.618) phalangeal proportions — critical for ML as it allows direct mapping of human mocap data without lossy retargeting. Weight: 1,300g including all sensing electronics.

DTA Tactile: 1,000+ pixels/fingertip, sub-mm spatial resolution, 6-axis force (Fx,Fy,Fz,Mx,My,Mz), 0.005N–30N range, 6,000 pressure levels, 180 FPS at 20ms latency. Survives 100K+ press cycles at 16N.

Comms: 1000Base-T Ethernet, 500Hz control. Sim: ROS 2, Isaac Gym/Lab, MuJoCo. Gesture speed: >4 Hz full open/close.

22 DoFProductionGolden RatioDTA Tactile6-axis Force

Wuji Hand — $5,500–$16,000

Wuji Robotics (Shanghai) — Masterclass in electromechanical actuation

20 DoF, all independently actuated. Abandons tendons entirely for self-locking rotary direct-drive micro-joint architecture. Motors, encoders, and gear reducers housed directly within fingers and palm. No mechanical coupling between joints — enables complex motions like piano playing. Skeleton: <600g.

IP Joints: Micro-worm gears (self-locking — grip holds indefinitely without power). MCP Joints: Four-bar linkage with palm motors delivering 3x finger torque.

Control: 1kHz FOC on all 20 axes, dual encoders (input + output) per joint. Power: 12-20V DC, 1A static / 35A peak. Comms: USB-C / RS485 / EtherCAT. SDK: wujihandpy (MIT license), ROS 2, URDF. Durability: 300K+ QC cycles, survives 80cm drop.

20 DoFProductionSelf-LockingWorm GearFour-BarNo Tendons

Shadow Dexterous Hand — $100,000+

Shadow Robot Company (UK) — Research gold standard since 2004

24 joints, 20 DoF. Tendon-driven with pneumatic muscle or electric motor options. 1:1 human scale. ~4,000g. >10N/tip force. The most widely used research dexterous hand in academia with hundreds of published papers.

Tactile: Optional BioTac (SynTouch) multimodal sensors. Comms: EtherCAT. Sim: ROS, Gazebo, MuJoCo. Extensive 3rd-party sim-to-real RL research base.

24 JointsProductionTendonResearch StandardBioTac

Psyonic Ability Hand — ~$10,000

Psyonic (USA) — FDA-cleared prosthetic, adaptable for robotics

6 DoF (6 independent actuators). Extremely fast: closes in 200ms. Originally designed as myoelectric prosthesis — world's fastest multi-articulating prosthetic. Built-in pressure sensors on each fingertip. Rugged construction for daily prosthetic use. Touch feedback via sensory substitution.

6 DoFProductionProstheticFDA Cleared200ms Close

Allegro Hand — ~$15,000

Wonik Robotics (South Korea) — Popular RL research platform

16 DoF (4 fingers x 4 joints). Torque-controlled via servo motors. 1.1 kg. Widely used in reinforcement learning and teleoperation research. Well-supported sim models in MuJoCo and Isaac Gym. Position, velocity, and torque control modes.

16 DoFProductionTorque ControlRL ResearchMuJoCo

Inspire Hand — ~$2,000–$5,000

Inspire Robots (China) — Cost-effective research/industrial hand

6-12 DoF depending on model (Dexterous / RH56 series). Servo-actuated with position feedback. Models from basic grippers to multi-finger dexterous variants. Used on multiple Chinese humanoid platforms. ROS support, affordable entry point for researchers.

6-12 DoFProductionServoBudgetROS

Robotiq Adaptive Grippers — $5,000–$15,000

Robotiq (Canada) — Industry-standard adaptive grippers

2-3 finger designs (2F-85, 2F-140, Hand-E). Not full dexterous hands but widely deployed adaptive grippers. Plug-and-play with Universal Robots. Force-controlled, position feedback. 5kg+ payload. Production-hardened with >10M cycle life.

2-3 DoFProductionIndustrialAdaptiveUR Compatible

Unitree Dex5-1 / Dex5-1P — $25,000–$29,000

Unitree Robotics (China) — First-party dexterous hand for H1/H2/G1

20 DoF (16 active + 4 passive coupled joints). Uses proprietary hollow-cup (coreless) motors with self-developed micro force-controlled composite transmission joints. 12 composite transmission joints + 4 gear transmission joints. All joints are smoothly backdrivable — a deliberate design choice for safe RL training. Micro-gap joint design places rotation axes close to finger surfaces, preventing object snagging.

Force: 10N/fingertip, 3.5kg grip (palm down) / 4.5kg (sideways). Weight: ~1,100g. Dims: 217x128x72mm. Control: 1kHz via USB 2.0. Voltage: 24-60V DC. Temp: -20C to 60C. Positioning: ±1mm fingertip repeat accuracy.

Dex5-1P (tactile variant, +$4K): 94 force sensors distributed across palm (10), fingertips (6 each), phalanges (6 each), finger roots (6 each). Detection: 10g–2.5kg range, 20kg max load. Reports both pressure and temperature.

Sim: MuJoCo (unitree_mujoco), Isaac Lab (unitree_sim_isaaclab), ROS/ROS 2 (URDF), Gazebo. SDK: unitree_sdk2 (C++), unitree_sdk2_python. Teleop: Apple Vision Pro, PICO 4, Meta Quest 3 via xr_teleoperate. Modular: Individual fingers replaceable.

Compatible: Unitree H1, H1-2, H2, G1 EDU (U1)+, and third-party via USB.

20 DoFProductionHollow-Cup MotorBackdrivable94 Sensors (P)1kHzIsaac LabMuJoCo

BrainCo Revo 2 — $5,000–$8,400

BrainCo (China/USA, Harvard Innovation Lab origin) — Prosthetic heritage, class-leading grip-to-weight

11 DoF (6 active + 5 passive). Only 383g — lightest production dexterous hand available. Despite minimal weight, delivers 50N grip force, 15N pinch force, and lifts 20kg single-hand — a grip-to-weight ratio of 52.6. Aerospace-grade aluminum frame with food-grade polymer contact surfaces. <50dB operation. 9-64V input range.

Three variants: Basic ($5K, RS-485 + CAN-FD) | Pro ($6.7K, + EtherCAT + adaptive impedance) | Touch ($7.5-8.4K, + 3D multimodal tactile: pressure, force direction, hardness, texture, proximity detection).

Precision: 0.1° repeatability, 0.1mm positioning. Speed: ≤0.65s full open-close. Control: ~1kHz. Comms: RS-485, CAN-FD, EtherCAT (Pro/Touch), Protobuf, DDS.

SDK: Python (3.8-3.12), C/C++, Linux/macOS/Windows. ROS 1 & ROS 2 (URDF via revo2_description, Apache 2.0). Unitree DDS bridge via brainco_hand_service.

BCI Heritage: BrainCo's prosthetic hand (TIME 100 Best Inventions 2019) uses 8-channel surface EMG for individual finger control. EMG armband teleoperation documented for Revo 2. Nature Communications publication on EEG-based BCI for real-time robotic hand control.

Compatible: Unitree H2, G1-D (Standard & Ultimate editions), R1 Edu Pro D, 6-DoF robotic arms, general platforms via flange interface.

11 DoFProduction383g20kg Lift52.6 G/W RatioEMG/BCIEtherCATROS 2

Company-Integrated Hands (Built into Humanoids)

Figure AI Hands

Custom direct-drive in-palm actuated hands with integrated sensing. Demonstrated folding shirts, sorting items, and delicate manipulation. Harmonic drive actuators for rotary joints. Undisclosed DoF but high dexterity shown in demos.

USADirect-DriveIntegrated

Tesla Optimus Hands

22 DoF (11 per hand). Wire/tendon-driven with actuators in forearm. Demonstrated threading needles, handling eggs, playing jenga. High sim-to-real gap due to tendon nonlinearities — active area of development with Grok integration.

USA22 DoFTendon

1X NEO Hands

Tendon-driven with actuators in forearm/torso. Biomimetic muscle routing for natural compliance. Designed for household tasks — soft, compliant interaction. World model approach (1XWM) for manipulation learning.

USATendonBiomimetic

Unitree Dex5-1 / Dex5-1P

20 DoF (16+4) dexterous hand for H1/H2/G1. Hollow-cup motors with micro gear transmission, fully backdrivable. $25K base / $29K with 94 tactile sensors. 1kHz control, MuJoCo + Isaac Lab + ROS 2. See detailed card above.

China20 DoF$25-29K94 Sensors

BrainCo Revo 2

11 DoF (6+5), only 383g but lifts 20kg (grip-to-weight 52.6). Three variants: Basic ($5K), Pro ($6.7K), Touch ($8.4K with 3D multimodal tactile). Harvard BCI heritage. Ships on Unitree H2, G1-D, R1. See detailed card above.

China/USA383g20kg LiftEMG/BCI

Sanctuary AI Phoenix Hand

Custom tendon-driven hand designed for general-purpose manipulation. 20+ DoF. Core IP of Sanctuary AI — designed specifically for their "Carbon" AI system. Demonstrated retail tasks at Mark's stores.

CanadaTendon20+ DoF

Clone Alpha Hands

Biomimetic musculoskeletal hands using McKibben-style pneumatic artificial muscles. Replicates human tendon routing. Highly compliant but difficult to control precisely. Unique approach in the field.

PolandPneumaticBiomimetic

Open-Source Dexterous Hands

Hand ↕InstitutionDoFActuationBOM ↕SimTactileLicenseKey Feature

Open-Source Hand Detail Cards

LEAP Hand v2 Advanced (CMU)

Software Ecosystem Leader — Deepak Pathak & Kenneth Shaw

20 DoF with hybrid rigid-soft architecture: 3D-printed rigid bones wrapped in compliant TPU exterior mimicking human skeletal structure. Powered palm articulations (one spanning four fingers, one near thumb) for natural palmar arch flexibility. BOM ~$5,080 with recommended CNC metal palm upgrade.

Software: Python/C++ APIs, tuned Isaac Gym/Lab environments, Apple Vision Pro and Manus glove teleoperation. Most mature open-source hand software stack available.

Supply chain risk: Dynamixel servo dependency creates risk for strict domestic manufacturing.

20 DoFDynamixelIsaac LabMuJoCoVR Teleop

RAPID Hand (Sun Yat-sen University)

Closest Wuji Analog — Direct-drive BLDC, perception-first

20 DoF, fully direct-driven. Physical architecture very similar to Wuji Hand. Primary contribution: co-optimized hardware-level perception framework that stably integrates wrist-mounted vision, fingertip tactile sensing, and proprioception with sub-7ms latency and perfect spatial alignment.

Perception integration has demonstrated superior performance in training diffusion policies compared to prior baselines. The most relevant open-source reference for replicating Wuji's actuation approach.

20 DoFDirect-Drive BLDCVBTSDiffusion PolicySub-7ms

ORCA Hand (ETH Zurich)

Durability Champion — Crash-proof poppable joints

17 DoF, tendon-driven with integrated FSR tactile sensors. Unique "popping joint" design: under excessive load, joints safely dislocate rather than breaking, protecting actuators. Survives 10,000+ continuous cycles (~20 hours non-stop manipulation) without hardware failure.

Auto-calibration via low-friction tendons routed through center of rotation. BOM under $2,200. Ideal for early RL training where crashes are frequent.

17 DoFTendonPoppable JointsFSR<$2,200

LEAP Hand v1 (CMU)

16 DoF. Direct-drive Dynamixel servos. BOM ~$2,000. MIT license. Strong simulation support with Isaac Lab and MuJoCo. Predecessor to v2 — simpler design, widely replicated. No tactile sensing in base design.

16 DoFDynamixelMIT~$2,000

RUKA Hand (NYU)

15 DoF. Tendon-driven with 11 Dynamixel servos. BOM $1,300. MuJoCo simulation support. Compact design optimized for research on in-hand manipulation. Open-source CAD and firmware.

15 DoFTendonDynamixel$1,300

Aero Hand Open

16 DoF (7 independently actuated). Tendon-driven. BOM only $314 — most affordable multi-finger hand. CC BY-SA license. MuJoCo support. Tendon force sensing. Excellent entry point for researchers on a budget.

16 DoFTendonCC BY-SA$314

BiDexHand

16+5 DoF (21 total). Cable-and-pulley actuation. BOM ~$400. MIT license. ROS 2 support. Built-in force sensors. Designed for bimanual dexterous manipulation research. Excellent value for dual-hand setups at ~$800 total.

21 DoFCable-PulleyMIT~$400

Wuji Architecture Deep Dive — The Self-Locking Advantage

IP Joints: Micro-Worm Gears

At each interphalangeal joint, the BLDC motor acts as the structural bone itself. A 90-degree micro-worm gear converts high-speed rotation into powerful flexion/extension. Worm gears are inherently self-locking: when power is cut, thread geometry prevents back-driving. The hand maintains massive grip force indefinitely without consuming power or generating heat.

For surgical instruments requiring 30-60 min continuous retraction, self-locking is transformative vs. standard planetary/direct-drive requiring continuous motor current.

MCP Joints: Four-Bar Linkages

At metacarpophalangeal joints (primary knuckles), a four-bar linkage mechanism allows a significantly larger, more powerful motor housed in the palm base, generating up to 3x the torque of finger joints. This dual architecture — worm gears for fingers, linkage-amplified motors for knuckles — explains how Wuji achieves 15N fingertip force and 10-20 kg full-hand grasp with human-scale motors.

Why Tendons Fail for Learned Manipulation

Tendon-driven hands (Shadow at $100K+, Tesla Optimus) suffer from a critical sim-to-real gap: cable tension varies dynamically with friction, stretching, temperature, and wear, making accurate simulation extremely difficult. Wuji's in-joint actuation eliminates these nonlinearities.

Note: Direct-drive reduces only the actuation modeling gap. Contact dynamics, sensor noise, environmental variation, and object properties still require domain randomization during policy training.

Critical Gap in Open-Source

No open-source true BLDC direct-drive hand with in-phalanx worm gears exists today.

The RAPID Hand is the closest analog, but replicating Wuji's self-locking worm gear mechanism at IP joints requires custom motor-transmission integration from scratch. This represents the key engineering challenge for any custom build.

Visuo-Tactile Sensing — Open-Source Platforms

Vision-Based Tactile Sensors (VBTS) — "feel by seeing" — use a clear silicone elastomer pad with reflective membrane, RGB LEDs, and miniature CMOS camera to produce 1,000+ tactile pixels matching camera resolution, achieving SharpaWave-level sensitivity without complex wiring.

SensorTypeCostSourceKey FeatureResolution

Transmission Systems Comparison

TransmissionBest ForAdvantagesDisadvantagesUS SourceCost/Joint

US BLDC Motor Supply Chain

Domestic suppliers for manufacturing dexterous hands without relying on foreign smart servos (Dynamixels) or Chinese micro-motors. The US hosts a sophisticated aerospace/medical motor base pivotable to humanoid robotics.

SupplierProductsRelevanceKey Advantage
Allied Motion / AllientPerformeX slotless BLDC, HT/Megaflux frameless torque motorsHand finger motors (15-20mm)2x performance of conventional motors at equivalent size; patented slotless design eliminates cogging
PortescapBrushless slotless mini motorsIP joint actuationOriginally designed for surgical power tools — withstands sterilization and continuous operation; ideal thermal management
KollmorgenFrameless servo motorsPalm-housed MCP motorsAerospace/satellite heritage; failure-proof operation; exceptional torque density with minimal EMI
TeknicClearPath integrated servosMCP four-bar linkage actuationBuilt-in encoders and drivers; slightly larger form factor ideal for palm housing
mjbotsmoteus-c1 controller ($69)FOC motor control for any BLDCOpen-source HW/FW; 38x38mm; CAN-FD; pairs with any BLDC motor; MIT license

Materials & Manufacturing

MaterialApplicationPropertiesUS Supplier
CF-Nylon (PA12+CF)Hand skeleton, phalangesCarbon fiber reinforced nylon; extreme rigidity + impact resistance; lighter than metal3DXTech (Michigan), MatterHackers (CA)
PEEKGear components, motor mountsMechanical properties rivaling aerospace aluminum; chemical resistance; 134C autoclave safe3DXTech (Michigan)
TPUCompliant finger exteriorsPassive rubber-like compliance for safe grasping; hybrid rigid-soft (proven by LEAP v2)MatterHackers (CA)
CF-Nylon SLS/MJFProduction structural partsNear-isotropic ±0.2mm tolerance; no support structures; complex pre-assembled linkagesProtolabs, RP America
Al 6061-T6Joint pins, actuator housings, wristGeneral-purpose CNC aluminum for structural componentsXometry, Protolabs
Ti Grade 5Surgical biocompatible componentsISO 10993-1 biocompatibility for surgical applicationsXometry, Protolabs

Build Configurations & BOM

Config A: Dynamixel Prototype — $4,560

Fast Path (5-6 months) — LEAP v2 Advanced architecture + GelSight fingertips + ORCA poppable joints

Dynamixel XC330 (x16) + XM430 (x4 MCP)$2,760
DIY GelSight fingertip sensors (x5)$300
MJF CF-Nylon structural parts$500
CNC aluminum joint pins + wrist$600
U2D2 + Power Hub + wiring$200
TPU compliant exteriors$50
Fasteners, tendons, misc$150
TOTAL (Prototype)$4,560
50-Unit Cost$3,060

Config B: Custom Wuji-Style — $6,550

Production Path (7-10 months) — US BLDC + worm gears (IP) + strain wave (MCP)

US/Chinese BLDC motors 12-18mm (x20)$1,500
Micro-worm gears IP joints (x16)$500
Strain wave gears MCP joints (x4)$600
AS5048A encoders input+output (x40)$200
Custom FOC driver PCBs TMC4671 (x20)$700
STM32H7 central controller$150
CNC + SLS CF-Nylon structural$2,200
DIY GelSight sensors (x5)$300
Bearings, fasteners, wiring, TPU$400
TOTAL (Prototype)$6,550
50-Unit Cost$2,900
NRE (one-time)$35-55K
Fully-loaded unit @50$3,600-4,000

Simulation & Sim-to-Real Pipeline

ComponentToolNotes
Training FrameworkNVIDIA Isaac Lab v2.3GPU-native, PyTorch tensor interface, automatic domain randomization
Physics EngineNVIDIA Newton / PhysX 5313x faster than MJX for manipulation tasks
Hand Sim ModelLEAP Hand Isaac Labgithub.com/leap-hand/LEAP_Hand_Isaac_Lab
Tactile SimulationTacSL (NVIDIA)GelSight in Isaac Lab, 200x faster than prior methods
Tactile Sim Alt.TACTO (Meta AI)Renders DIGIT output in physics engines at 100s FPS
Policy Learning3D-ViTac methodologyFuses visual + tactile into unified 3D point cloud for diffusion policy
Surgical TasksORBIT-Surgical14 benchmark tasks: needle grasp, suture, handover
TeleoperationApple Vision Pro / Manus glovesDemonstration collection for imitation learning
Motor ControlSimpleFOC + STM32G4Open-source FOC framework; dual AS5048A encoders per joint
CommunicationCAN-FD (5 Mbps)20 motors @1kHz = 2.048 Mbps; ~$1/node via MCP2562FD

Phased Actuator Strategy

PhaseActuatorPer-Unit20-Axis TotalForceSim GapNotes
Phase 1Dynamixel XC330~$90$1,8008-12NModerateFastest start; LEAP ecosystem works out-of-box
Phase 1+Dynamixel XM430 (MCP)~$3304x = $1,320>20NModerateForce upgrade for MCP joints only
Phase 2US BLDC + worm gear~$150-300$3,000-6,00015-25NLowSelf-locking, surgical-grade, domestic supply
BudgetChinese BLDC + moteus~$100-150$2,000-3,00015-25NLowBest value for QDD path

Key Repositories & Resources

LEAP Hand v2 Advanced

github.com/leap-hand — CAD, API, Isaac Lab envs, teleoperation

DIGIT Tactile Sensor

github.com/facebookresearch/digit-design — PCB layouts, firmware, 3D models, BOM

TACTO Simulator

github.com/facebookresearch/tacto — Vision-based tactile simulation in PyBullet

ORCA Hand

orca.ethz.ch — Poppable joint design, auto-calibration

RAPID Hand

arxiv.org/abs/2506.07490 — 20-DoF direct-drive, perception framework

3D-ViTac

binghao-huang.github.io/3D-ViTac — Unified visuo-tactile policy learning

SimpleFOC

simplefoc.com — Open-source FOC motor control library

moteus Controller

github.com/mjbots/moteus — Open-source HW/FW BLDC servo controller ($69)

ORBIT-Surgical

orbit-surgical.github.io — 14 benchmark surgical tasks in Isaac Lab

3D-Printable Strain Waves

hackaday.io/project/165599 — 30:1 ratio harmonic drives, <$20/joint

NSF HAND Center

Northwestern, CMU, MIT, Florida A&M, Texas A&M — $26M+ center for industry collaboration. PI: Ed Colgate / Kevin Lynch

AnySkin (NYU)

any-skin.github.io — Plug-and-play replaceable magnetic tactile skins

Cost Comparison Summary

Shadow Dexterous Hand
$100,000+
SharpaWave
$50,000
Unitree Dex5-1P
$29,000
Unitree Dex5-1
$25,000
Wuji Hand (high)
$16,000
Allegro Hand
$15,000
Psyonic Ability
$10,000
BrainCo Revo 2 Touch
$8,400
BrainCo Revo 2 Pro
$6,700
Config B (prototype)
$6,550
Wuji Hand (low)
$5,500
LEAP v2 Adv BOM
$5,080
BrainCo Revo 2 Basic
$5,000
Config A (prototype)
$4,560
Config B @50 units
$2,900
ORCA Hand BOM
$2,200
LEAP v1 BOM
$2,000
RUKA BOM
$1,300
BiDexHand BOM
$400
Aero Hand BOM
$314

Gears & Transmission Types by Company

Dominant or most likely gear/actuation families per vendor. Confidence: High = public disclosure; Medium = inferred from teardowns/patents; Low = best-read assumption.

CompanyGear / Actuation TypeConfidence

Embodied AI: Two Power Blocs

LayerTypical PlayersStrategy
"Brains" — Multimodal LLMsOpenAI, xAI, Google DeepMind, Anthropic, NVIDIA, Meta, MicrosoftScale frontier models; retune from text-only to VLA control; license/open-source interfaces for robot makers
"Bodies" — Humanoid OEMsTesla, Figure AI, Agility, Apptronik, 1X, Sanctuary AI, Fourier, Unitree, UBTECHProve reliable mechatronics; raise massive rounds; sign first commercial pilots

LLM Vendors in Embodied AI

OpenAI

Led 1X Series A & Figure Series B. Training GPT-4o-Vision + Voice on humanoid sensor data.

1XFigure AI

xAI / Tesla

Grok 4 co-developed with Optimus Gen-3. Vision-language-action control into real-time sensor fusion.

Tesla Optimus340K GPU cluster

Google DeepMind

RT-2 → RT-X → Gemini Robotics-ER (2025). Partnered with Apptronik for humanoid trials.

ApptronikOpen VLA

NVIDIA

GR00T N1 open VLA model on Fourier GR-1. Pull-through for Thor & Jetson hardware. Isaac Sim ecosystem.

FourierUnitreeAgility1X

Microsoft

Azure Robotics Platform. Invested in Figure ($675M round) and Sanctuary research MoU.

FigureSanctuary

Meta (LLaMA)

Reality Labs Robotics division. LLaMA-3 VLA models. Early talks with Unitree & Figure.

UnitreeFigure

1X World Model vs Figure Helix

Axis1X World Model (1XWM)Figure Helix
PurposeGenerative simulator ("dreaming engine") — predicts future sensor framesReal-time VLA policy — converts instruction to 200Hz whole-body control
ArchitectureDiffusion-based video generator + action conditioningSystem-2 (7B VLM @ 7-9Hz) + System-1 (80M visuomotor @ 200Hz)
Data1000s hrs on-board RGB-D, joint states, proprioceptive in homes/offices~500 hrs teleop data achieved broad generalisation
LatencyOffline (GPU-intensive)5-10ms end-to-end on-device
Key AdvantageGround-truth physics; synthetic data flywheel for edge casesOn-device privacy; meets ISO 10218-2; runs under 50W

Scenarios to 2030

Layered Equilibrium (Most Likely)

LLM vendors publish certified "robot brains" APIs; OEMs compete on price, dexterity, and vertical-market integrations.

LLM Vendor Sweep

One or two hyperscalers bundle compute + model + reference robot and commoditise the rest.

Hardware Resurgence

Breakthrough in cheap actuators lets OEMs detach from cloud models and train small on-device VLAs.

Regulatory Stall-out

Dual-use export controls choke cross-border supply chains; adoption tilts to regulated domestic pilots only.

Safety Standards & Certification

StandardDomainRelevance to Humanoids
ISO 10218-1/-2IndustrialMachinery safety for articulated industrial robots
ISO/TS 15066CollaborativePermissible contact pressure for human-robot collaboration
IEC 61508 / ISO 13849Functional SafetySIL/PL targets for control systems (torque-limited joints, brakes)
UL 4600 Ed. 3AutonomySafety case for autonomous mobile operations
ISO 13482:2014Personal CareForce/speed caps for personal-care robots in homes
IEC 60335-1Domestic ElectricalClearances, creepage, over-current for household devices
ISO/WD 25785-1Mobile IndustrialSafety for industrial mobile robots with actively controlled stability
ISO/IEC 27001CybersecurityCertifiable framework for corporate security policies
IEC 62443ICS SecurityCybersecurity for industrial automation control systems
GDPRData PrivacyConsent, data minimization, encryption for robot-collected data
EU AI ActAI RegulationHigh-risk classification; risk management, transparency, human oversight
NIST CSFCyber FrameworkIdentify, Protect, Detect, Respond, Recover for robot deployments

Safety Engineering Pillars

Mechanical Limits

Lightweight CFRP skeleton, joint travel hard-stops, energy-absorbing skins. ISO 12100.

Force & Power Limitation

Series-elastic actuators, velocity & torque monitoring, zero-power brakes. ISO/TS 15066.

Redundant Perception

Multi-modal depth cameras + LiDAR for 360° occupancy; tactile skins. UL 4600.

Electrical Safety

Double insulation, PELV 48 VDC buses, surge suppression. IEC 60204-1.

Cybersecurity

Secure boot, OTA signing, GDPR-compliant data handling. ISO/IEC 27001.

Human Factors

Predictable motion, intuitive intent signalling (LED, speech), teachable modes. ISO 9241-210.

Key Cyber Threats

Remote Hijacking

Attackers gain control via network or malicious firmware. Demonstrated on SoftBank NAO.

Data Leakage

Cameras/mics coerced into capturing personal data. Unauthorized sensor access.

Behavior Manipulation

Poisoned training data makes robots learn unsafe behaviors.

Adversarial Inputs

Crafted visual/audio inputs fool perception systems into misclassifying objects.

The U.S.-China Contest Over Humanoids & Embodied AI

China's State-led Sprint

Policy LeverEvidenceIntended Effect
Industrial PlanMIIT guideline: "realise mass production by 2025"Fast-track R&D, rally local governments & SOEs
Capital Mobilisation83 funding events in H1-2025 raised ≈¥6.9B (~$950M)Flood sector with patient money
Champions & IPOsUnitree preparing 2025 domestic IPOAnchor supply chains and talent
Domestic ComputeHuawei Ascend 910 chips; courting buyers abroadReduce dependence on Nvidia/AMD

U.S. Multi-track Response

TrackActionStrategic Aim
Export ControlsNew BIS rules (Apr 2025) stop Nvidia H20 to China (~$8B revenue hit)Deny China most efficient training/inference silicon
Outbound InvestmentTreasury rule under EO 14105 restricts U.S. capital in China's AIPrevent "know-how by cheque"
Domestic IncentivesCHIPS Act + DoD AI Rapid Capabilities Cell (Nov 2024)Keep advanced pipelines onshore
Allied Market AccessBIS eases restrictions for trusted partnersLock U.S.-centric supply chain

Likely Scenarios 2025-2030

Bifurcated Stacks

Two incompatible HW/SW ecosystems; export bans expand to servomotors and Li-ion packs.

Managed Interdependence

U.S. chips in Chinese humanoids under license caps; JVs in neutral markets (Mideast logistics).

Throttled Innovation Spiral

Tit-for-tat curbs hamper global supply, slowing overall adoption; EU/Japan becomes "swing supplier."

Selective Convergence

Both powers relax on civilian lines (elder-care, retail) while cordoning off military uses.

Public Equities

#CompanyHumanoid ProjectTickerKey Notes
1Tesla Inc.Optimus production robotsNASDAQ: TSLAHighest-volume ticker; targeting 1000s of units in 2025
2Xiaomi Corp.CyberOne (Robotics Lab)HKEX: 1810Low entry price, deep trading pool; consumer-electronics synergies
3Hyundai Motor80% owner of Boston DynamicsKorea SE: 005380Only pure industrial play with direct BD stake
4SoftBank GroupPepper v2 & Vision Fund stakesOTC: SFTBY / Tokyo: 9984Diversified holding; liquidity best via ADR
5Toyota MotorT-HR3 & "Musashi" R&DNYSE: TMPartner-robot platform for healthcare & mobility

Pre-IPO (Available on Hiive)

Figure AI

Figure 01 general-purpose bot. Highest flow on Hiive among robotics names.

MicrosoftOpenAINvidiaBezosARK

1X Technologies

NEO home-service bot. OpenAI-backed.

NVIDIAEQTOpenAITiger GlobalSamsung NEXT

Sanctuary AI

Phoenix dexterous workspace bot. Canadian leader; regular secondary blocks.

AccentureMagnaVerizonWorkday

Apptronik

Apollo manufacturing assistant. Mercedes-Benz pilot in 2025.

GoogleMercedes-BenzARKJapan Post

Agility Robotics

Digit warehouse & logistics bot. Frequent bids/asks; recently raised Series C.

DCVCAmazonSonyTDK

Open-Source Humanoid Platforms

VendorPlatformPriceLicenseGitHub StarsDiscordUnits/Field2025 Milestones
Pollen Robotics (🇫🇷)Reachy 2€29K; 100+ shippedGPL v3 / CERN-OHL2,1004,000100+CES 2025 launch; Hugging Face acquisition (Apr 2025)
K-Scale Labs (🇺🇸)K-Bot / Z-Bot$8,999; Nov 2025MIT + CERN-OHL-S1442,300Pre-ordersFirst YC-backed humanoid; full RL locomotion + VLA alpha
UC Berkeley (🇺🇸)Berkeley Humanoid Lite<$5K self-buildBSD-3-Clause8201,100200+ BOM downloadsRSS 2025: zero-shot sim-to-real walking + cube manipulation
InMoov CommunityInMoov (1.8m)~$1,500 DIYCC-BY-NC5,80010,000+ partialWorld's largest DIY humanoid community
Poppy Project (🇫🇷)Poppy Humanoid€8K kitGPL v3 / CC-BY-SA60+ research labsROS 2 port Feb 2025; €1.6M EU Horizon grant
NimbRo (Univ. Bonn 🇩🇪)NimbRo-OP2X€35K researchGPL v3ResearchAGILOped redesign for RoboCup AdultSize

Key Technical Trends

Simulation-First Development

All vendors ship MuJoCo or Isaac Gym environments. Millions of control iterations before hardware moves.

Open Foundation "Brains"

NVIDIA GR00T N1 (open weights, Mar 2025) and Hugging Face twin humanoids (May 2025).

ROS 2 + Rust Runtimes

ROS 2 Foxy/Humble messaging backbone. Real-time control migrating to Rust for safety (K-Scale K-OS).

Licensing: GPL v3 + CERN-OHL-S

Dual-stack copyleft pattern dominates. BSD/MIT for academic spin-outs.

Key Milestones in Humanoid Robotics

2000
Honda unveils ASIMO — First self-contained biped able to walk stairs, proving dynamic balance
2011
ASIMO Gen 2 — Adds running and multi-finger dexterity, inspiring global R&D push
2013-15
Boston Dynamics Atlas prototypes — Rough-terrain walking and tool use under DARPA funding
2017
Atlas back-flip & parkour demo — Human-level power-to-weight and control; video goes viral
2019-20
Agility Robotics Digit ships to Ford — First legged robot sold for commercial pilot (package delivery)
2021
Tesla announces Optimus — Signals mega-cap entry; targets sub-$20K price
2023
Optimus Gen 2 prototype — Fine-motor palletizing; long-term cost goal reiterated
2024
1X Technologies debuts NEO — AI-centric "everyday companion" with tendon-driven arms
2024
Unitree G1 launched at $5,600 — Cost-curve compression: 99.4% drop from Atlas research unit
2025
Digit hired by GXO Logistics — First multi-year paid RaaS contract in live warehouse ops
2025
Figure AI Series C — $1.5B raise at ~$40B valuation; Nvidia, SoftBank, Tiger Global
2025
Grok 4 × Optimus — First frontier-LLM co-trained on live humanoid sensor logs
2026
Unitree, UBTECH, EngineAI — Multi-1000 delivery contracts signed. Humanoids deployed in Chinese industry
2026
1X, Figure, Tesla — Ramp manufacturing for early adopters in industry and domestic

S-Curve Adoption Phases

PhaseUnit VolumesPrice Band (USD)Typical BuyerSuccess Metric
2025-27 Pilot EraHundreds → low thousands$50-150KEnterprise pilotsReliability ≥ 95%
2028-32 Early ScaleTens of thousands/yr$20-50KLarge-fleet industrial + premium homes2-yr payback
2033+ Mass UptakeMillions/yr< $15KMainstream SMEs & householdsHuman-parity dexterity

Decade Roadmap

2025-26

Pilot fleets (>1,000 units); 95% uptime; ISO draft standard.

2027-29

BoM ≈ $10K; consumer beta units for cleaning & laundry.

2030-32

BoM < $5K; multi-million installed base; services > hardware revenue.

2033+

Human-parity dexterity; solid-state batteries; trillion-dollar services economy.

Five Overarching Takeaways

1. Cost Curve is Collapsing

Atlas-era $1M robots → sub-$6K production units. >90% drop in 6 years; $20K feasible before 2030.

2. Industrial Pilots Pay the R&D Bill

Logistics contracts already fund multi-thousand-unit pipelines. Consumer volumes lag ~5 years.

3. Consumers Crave Drudgery Relief

Cleaning, laundry, and garden chores top every wish-list. Companionship is the only non-chore request.

4. Hands Matter More Than Hype

98% grasp reliability outranks back-flip videos. Manipulation mastery is the king-maker capability.

5. Standards Are the Next Moat

Early movers in safety and privacy regulation will lock in advantage. Ecosystem consolidation inevitable.

Stakeholder Playbook

StakeholderDo Now (2025-27)Prepare For (2028-32)
InvestorsDiversify across brain-body-integrator; insist on ≤24-mo pilot paybackDouble down on emerging winners; safe + capable home robots
Robot OEMsNail one paid task at 99% uptime; lock actuator supplyScale volume; open app store; drive BoM < $20K
Component ProvidersCo-design standard actuators/sensors with top OEMsOffer integrated sensor bundles
AI/Software ProvidersCollaborate on perception stack and task autonomyLicense perception-as-a-service; integrate cloud & edge
Enterprise AdoptersRun 3-6 mo pilots; capture downtime & safety KPIsNegotiate fleet leases; integrate with WMS/ERP
Policy-MakersDraft baseline mobile-service-robot rules; fund re-skillingEstablish certification labs; monitor labour displacement

Watch-List — Early Warning Signals

Positive TriggerNegative Trigger
Sub-$30K retail from Tier-1 OEMHigh-profile home accident → moratorium
First pure-play humanoid IPOActuator/battery supply crunch
ISO humanoid safety standard issuedInsurance premiums spike after injury lawsuit
Multi-modal AI masters reliable folding & cookingHome robot privacy breach
National "robot driver-licence" framework passesActuator supply crunch delays shipments >12m