[{"data":1,"prerenderedAt":874},["ShallowReactive",2],{"glossary-list-en":3},[4,93,166,236,325,390,488,565,640,708,786],{"id":5,"title":6,"alternateName":7,"body":8,"description":83,"extension":84,"keywords":85,"meta":86,"navigation":87,"path":88,"seo":89,"stem":90,"updated":91,"__hash__":92},"glossary\u002Fglossary\u002Fen\u002Fcrossed-roller-bearing.md","Crossed Roller Bearing","交叉滚子轴承",{"type":9,"value":10,"toc":78},"minimark",[11,16,25,30,37,70],[12,13,15],"h1",{"id":14},"what-is-a-crossed-roller-bearing","What Is a Crossed Roller Bearing?",[17,18,19,20,24],"p",{},"A ",[21,22,23],"strong",{},"crossed roller bearing"," is a precision bearing whose cylindrical rollers alternate orientation at 90° to each other inside a V-shaped raceway. Because adjacent rollers face opposite directions, one bearing simultaneously carries radial loads, bidirectional axial loads, and overturning (moment) loads.",[26,27,29],"h2",{"id":28},"why-robot-joints-prefer-it","Why Robot Joints Prefer It",[17,31,32,33,36],{},"Standard deep-groove ball bearings mainly take radial load; covering combined loads usually requires paired bearings and a bigger structure. Robot joint loads are inherently combined — a humanoid hip joint simultaneously supports body weight (axial), leg-swing centrifugal force (radial), and the bending moment of a cantilevered leg (overturning). A crossed roller bearing handles ",[21,34,35],{},"all three load types in a single bearing position",":",[38,39,40,47,59],"ul",{},[41,42,43,46],"li",{},[21,44,45],{},"High stiffness",": roller line contact resists deformation better than ball point contact.",[41,48,49,52,53,58],{},[21,50,51],{},"High rotational precision",": a stable reference for output-side measurement by ",[54,55,57],"a",{"href":56},"\u002Fen\u002Fglossary\u002Fdual-absolute-encoder","dual absolute encoders",".",[41,60,61,64,65,69],{},[21,62,63],{},"Space savings",": the single-bearing design keeps ",[54,66,68],{"href":67},"\u002Fen\u002Fglossary\u002Fjoint-motor","joint motors"," thinner and lighter.",[17,71,72,73,77],{},"The BXI ",[54,74,76],{"href":75},"\u002Fen\u002Fmotors\u002Fadvanced-motors","85\u002F70\u002F50-series joint motors"," fit crossed roller bearings at the output across the whole lineup, from load-bearing legs to dexterous arms.",{"title":79,"searchDepth":80,"depth":80,"links":81},"",2,[82],{"id":28,"depth":80,"text":29},"A crossed roller bearing alternates cylindrical rollers at 90° in a V-shaped raceway, letting a single bearing carry radial, axial, and moment loads simultaneously — giving robot joints high stiffness and rotational precision.","md","crossed roller bearing, robot bearing, joint stiffness, moment load",{},true,"\u002Fglossary\u002Fen\u002Fcrossed-roller-bearing",{"title":6,"description":83},"glossary\u002Fen\u002Fcrossed-roller-bearing",null,"KoMUdgdok7QNxJjSWGk9TwhPtuw-cktDynA8ZmwxPIU",{"id":94,"title":95,"alternateName":96,"body":97,"description":159,"extension":84,"keywords":160,"meta":161,"navigation":87,"path":162,"seo":163,"stem":164,"updated":91,"__hash__":165},"glossary\u002Fglossary\u002Fen\u002Fdegrees-of-freedom.md","Degrees of Freedom (DoF)","自由度",{"type":9,"value":98,"toc":155},[99,103,112,116,136,140,147],[12,100,102],{"id":101},"what-are-degrees-of-freedom","What Are Degrees of Freedom?",[17,104,105,108,109,58],{},[21,106,107],{},"Degrees of freedom (DoF)"," count the independently controllable motion axes of a robot. Each actively driven rotary or prismatic joint counts as one DoF, typically corresponding to one ",[54,110,111],{"href":67},"joint motor",[26,113,115],{"id":114},"dof-determines-capability","DoF Determines Capability",[38,117,118,124,130],{},[41,119,120,123],{},[21,121,122],{},"6 DoF"," is the minimum for a robot arm to reach \"any position + any orientation\" in 3D space (3 for position + 3 for orientation);",[41,125,126,129],{},[21,127,128],{},"7 DoF"," adds a redundant axis, letting the elbow move while the hand stays fixed — dodging obstacles and optimizing posture, exactly the configuration of the human arm;",[41,131,132,135],{},[21,133,134],{},"Humanoid robots"," need whole-body coordination and typically exceed 30 DoF.",[26,137,139],{"id":138},"example-the-31-dof-of-elf-3","Example: The 31 DoF of Elf 3",[17,141,72,142,146],{},[54,143,145],{"href":144},"\u002Fen\u002Frobots\u002Fhumanoid-robot","Elf 3 humanoid robot"," has 31 DoF excluding hands: 6 per leg, 7 per arm, 3 in the waist, 2 in the head. The 6-DoF legs cover walking's hip (3), knee (1), and ankle (2); the 7-DoF arms provide human-like manipulation redundancy; the 3-DoF waist expands the reachable workspace.",[17,148,149,150,154],{},"More DoF means more actuators, more control-bus bandwidth (see ",[54,151,153],{"href":152},"\u002Fen\u002Fglossary\u002Fmit-protocol-can","MIT protocol","), and more weight — making DoF layout one of the central trade-offs in humanoid design.",{"title":79,"searchDepth":80,"depth":80,"links":156},[157,158],{"id":114,"depth":80,"text":115},{"id":138,"depth":80,"text":139},"Degrees of freedom count the independently controllable joint axes of a robot and directly determine its motion capability: a 6-DoF arm reaches arbitrary poses, while humanoid robots typically need 30+ DoF for whole-body coordination.","degrees of freedom, DoF, robot DoF, humanoid robot DoF, redundant DoF",{},"\u002Fglossary\u002Fen\u002Fdegrees-of-freedom",{"title":95,"description":159},"glossary\u002Fen\u002Fdegrees-of-freedom","SkapjgIm8JHQbbVFUk-AwsALlHdvJjGwWAR3WQh8jaY",{"id":167,"title":168,"alternateName":169,"body":170,"description":229,"extension":84,"keywords":230,"meta":231,"navigation":87,"path":232,"seo":233,"stem":234,"updated":91,"__hash__":235},"glossary\u002Fglossary\u002Fen\u002Fdual-absolute-encoder.md","Dual Absolute Encoder","双绝对值编码器",{"type":9,"value":171,"toc":225},[172,176,185,189,196,216,220],[12,173,175],{"id":174},"what-is-a-dual-absolute-encoder","What Is a Dual Absolute Encoder?",[17,177,19,178,181,182,184],{},[21,179,180],{},"dual absolute encoder"," configuration is the encoder architecture of high-end ",[54,183,68],{"href":67},": one absolute encoder on the motor rotor (input side, before reduction) and one on the output flange (output side, after reduction), measuring angle directly at both ends.",[26,186,188],{"id":187},"why-two-encoders","Why Two Encoders?",[17,190,191,192,195],{},"With only a motor-side encoder, output angle must be inferred as \"motor angle ÷ reduction ratio\" — and gearbox backlash, elastic deformation, and assembly tolerance all pull that estimate away from the true joint angle. An output-side encoder ",[21,193,194],{},"measures the real post-reduction angle directly",", so the control loop closes on ground truth:",[38,197,198,204,210],{},[41,199,200,203],{},[21,201,202],{},"Higher precision",": angle errors from backlash and deformation are eliminated.",[41,205,206,209],{},[21,207,208],{},"Zero-calibration startup",": absolute encoders retain position through power loss, so the robot knows every joint pose at power-on with no homing routine.",[41,211,212,215],{},[21,213,214],{},"Redundancy",": input and output angles cross-check each other, catching anomalies such as gearbox slip.",[26,217,219],{"id":218},"implementation","Implementation",[17,221,72,222,224],{},[54,223,76],{"href":75}," pair a magnetic encoder on the input with an inductive encoder on the output — two sensing principles that don't interfere — achieving true dual-encoder closed loop in a compact package.",{"title":79,"searchDepth":80,"depth":80,"links":226},[227,228],{"id":187,"depth":80,"text":188},{"id":218,"depth":80,"text":219},"A dual absolute encoder setup places one absolute encoder at the joint motor input and another at the output, directly measuring true post-reduction joint angle for precise closed-loop control and zero-calibration startup.","dual absolute encoder, absolute encoder, joint angle sensing, zero calibration startup",{},"\u002Fglossary\u002Fen\u002Fdual-absolute-encoder",{"title":168,"description":229},"glossary\u002Fen\u002Fdual-absolute-encoder","G9sod0p6BHMaU_XSx7PUCyIPIdnjXMM1pHPzOdRf-WM",{"id":237,"title":238,"alternateName":239,"body":240,"description":318,"extension":84,"keywords":319,"meta":320,"navigation":87,"path":321,"seo":322,"stem":323,"updated":91,"__hash__":324},"glossary\u002Fglossary\u002Fen\u002Fembodied-ai.md","Embodied AI","具身智能 \u002F Embodied Intelligence",{"type":9,"value":241,"toc":314},[242,246,251,255,267,271,309],[12,243,245],{"id":244},"what-is-embodied-ai","What Is Embodied AI?",[17,247,248,250],{},[21,249,238],{}," (embodied intelligence) is the paradigm in which an intelligent agent has a physical body and accomplishes perception, decision-making, and action through interaction with the real environment. Unlike \"disembodied\" models that only process text or images, embodied AI must close the loop in the physical world: see → understand → act → observe the result → correct.",[26,252,254],{"id":253},"why-humanoids-are-the-primary-platform","Why Humanoids Are the Primary Platform",[17,256,257,258,261,262,266],{},"Human environments — stairs, door handles, tools, workstations — are designed for the human body. The humanoid form lets robots reuse this infrastructure ",[21,259,260],{},"without modifying the environment",", and lets vast human motion data (video, motion capture, ",[54,263,265],{"href":264},"\u002Fen\u002Fglossary\u002Fteleoperation","teleoperation"," demonstrations) transfer directly into training data.",[26,268,270],{"id":269},"the-stack","The Stack",[38,272,273,287,293,303],{},[41,274,275,278,279,281,282,286],{},[21,276,277],{},"Body",": high-dynamic hardware is the prerequisite — torque-dense ",[54,280,68],{"href":67},", high-frequency control buses, and whole-body ",[54,283,285],{"href":284},"\u002Fen\u002Fglossary\u002Fdegrees-of-freedom","degrees of freedom",";",[41,288,289,292],{},[21,290,291],{},"Perception",": multimodal sensing — vision, depth, IMU, touch;",[41,294,295,298,299,286],{},[21,296,297],{},"Decision",": foundation-model or reinforcement-learning policies, often transferred from simulation via ",[54,300,302],{"href":301},"\u002Fen\u002Fglossary\u002Fsim-to-real","Sim-to-Real",[41,304,305,308],{},[21,306,307],{},"Data",": teleoperated real-robot data collection for imitation learning.",[17,310,72,311,313],{},[54,312,145],{"href":144}," ships with a ROS2 SDK and MuJoCo simulation — an open platform built for embodied-AI research.",{"title":79,"searchDepth":80,"depth":80,"links":315},[316,317],{"id":253,"depth":80,"text":254},{"id":269,"depth":80,"text":270},"Embodied AI is the paradigm where an agent perceives, decides, and acts through a physical body interacting with the real world; humanoid robots are its primary platform, closing the perception–decision–action loop in reality.","embodied AI, embodied intelligence, humanoid robot, perception decision action, agents",{},"\u002Fglossary\u002Fen\u002Fembodied-ai",{"title":238,"description":318},"glossary\u002Fen\u002Fembodied-ai","C6NRbR956x72lE_j5DM_QaS5e_lb5_lRYo0H3qL44lY",{"id":326,"title":327,"alternateName":328,"body":329,"description":383,"extension":84,"keywords":384,"meta":385,"navigation":87,"path":386,"seo":387,"stem":388,"updated":91,"__hash__":389},"glossary\u002Fglossary\u002Fen\u002Fhollow-shaft-motor.md","Hollow Shaft Motor","中空轴电机",{"type":9,"value":330,"toc":379},[331,335,341,345,348,359,364,368],[12,332,334],{"id":333},"what-is-a-hollow-shaft-motor","What Is a Hollow Shaft Motor?",[17,336,19,337,340],{},[21,338,339],{},"hollow shaft motor"," is a motor or actuator whose output shaft has a through-bore along its rotation axis. Unlike solid-shaft designs, cables, hydraulic lines, and sensor harnesses can pass straight through the joint's center of rotation instead of looping around the outside.",[26,342,344],{"id":343},"why-robots-need-hollow-shafts","Why Robots Need Hollow Shafts",[17,346,347],{},"High-DOF robots — such as a 31-DOF humanoid — chain many joints in series. If every joint routes its cables externally:",[38,349,350,353,356],{},[41,351,352],{},"harnesses fatigue and fail where joints flex repeatedly;",[41,354,355],{},"exposed cables limit range of motion and snag on the environment;",[41,357,358],{},"wiring complexity and assembly time grow rapidly with joint count.",[17,360,361,362,58],{},"Routing through the center bore solves all three at once, which is why hollow-shaft designs are central to modern integrated ",[54,363,68],{"href":67},[26,365,367],{"id":366},"engineering-trade-offs","Engineering Trade-offs",[17,369,370,371,373,374,378],{},"A larger bore passes more wiring but complicates the layout of the rotor, encoders, and gearbox. The BXI ",[54,372,76],{"href":75}," provide 6–10 mm hollow bores, using a ",[54,375,377],{"href":376},"\u002Fen\u002Fglossary\u002Fplanetary-gearbox","planetary gearbox"," to balance torque density against routing space.",{"title":79,"searchDepth":80,"depth":80,"links":380},[381,382],{"id":343,"depth":80,"text":344},{"id":366,"depth":80,"text":367},"A hollow shaft motor has a through-bore along its output axis, letting cables, tubing, and sensor harnesses pass through the joint center — dramatically simplifying robot wiring and expanding joint range of motion.","hollow shaft motor, hollow bore actuator, robot cable routing, joint motor design",{},"\u002Fglossary\u002Fen\u002Fhollow-shaft-motor",{"title":327,"description":383},"glossary\u002Fen\u002Fhollow-shaft-motor","8AXKSNUXkdra2vm9kBh5SEf95zF0eH7tQEh9y--ROE4",{"id":391,"title":392,"alternateName":393,"body":394,"description":481,"extension":84,"keywords":482,"meta":483,"navigation":87,"path":484,"seo":485,"stem":486,"updated":91,"__hash__":487},"glossary\u002Fglossary\u002Fen\u002Fjoint-motor.md","Joint Motor","关节电机 \u002F Robot Actuator",{"type":9,"value":395,"toc":477},[396,400,405,409,452,456,470],[12,397,399],{"id":398},"what-is-a-joint-motor","What Is a Joint Motor?",[17,401,19,402,404],{},[21,403,111],{}," (also called a robot joint module or integrated actuator) packs a frameless torque motor, gearbox, encoders, driver, and bearings into one compact unit. Mounted directly at a robot's joints, it outputs controlled torque and rotation — the \"muscle\" of humanoid robots, quadrupeds, and robotic arms.",[26,406,408],{"id":407},"core-components","Core Components",[38,410,411,417,426,434,443],{},[41,412,413,416],{},[21,414,415],{},"Frameless torque motor",": generates raw torque and sets the power ceiling.",[41,418,419,422,423,425],{},[21,420,421],{},"Gearbox",": trades speed for torque; common choices are the ",[54,424,377],{"href":376}," and harmonic drive.",[41,427,428,431,432,58],{},[21,429,430],{},"Encoders",": measure joint angle for closed-loop control; high-end designs use ",[54,433,57],{"href":56},[41,435,436,439,440,58],{},[21,437,438],{},"Bearing",": carries loads while preserving rotational precision; load-bearing joints typically use a ",[54,441,23],{"href":442},"\u002Fen\u002Fglossary\u002Fcrossed-roller-bearing",[41,444,445,448,449,451],{},[21,446,447],{},"Driver",": runs current\u002Fvelocity\u002Fposition loops and receives commands over CAN\u002FCANFD buses (see ",[54,450,153],{"href":152},").",[26,453,455],{"id":454},"key-specifications","Key Specifications",[17,457,458,459,462,463,466,467,469],{},"The parameters that matter most are ",[21,460,461],{},"rated torque"," (continuous output), ",[21,464,465],{},"peak torque"," (short-term ceiling), weight, envelope dimensions, reduction ratio, and communication interface. A humanoid robot typically needs 20–40 joint motors across several torque tiers — the BXI Elf 3, for example, runs 31 ",[54,468,76],{"href":75}," spanning 25–150 N·m peak torque.",[17,471,472,473,58],{},"For a sizing methodology, see the ",[54,474,476],{"href":475},"\u002Fen\u002Fblog\u002Fjoint-motor-selection-guide","joint motor selection guide",{"title":79,"searchDepth":80,"depth":80,"links":478},[479,480],{"id":407,"depth":80,"text":408},{"id":454,"depth":80,"text":455},"A joint motor is an integrated robot actuator combining a frameless motor, gearbox, encoders, and driver in one unit, mounted directly at a robot joint — the core power component of humanoid robots and robotic arms.","joint motor, robot actuator, integrated actuator, robot joint module",{},"\u002Fglossary\u002Fen\u002Fjoint-motor",{"title":392,"description":481},"glossary\u002Fen\u002Fjoint-motor","x9EEPuA3LvSXMHvXgBzkE8Esmeqaip4sfS4pTLlHaL8",{"id":489,"title":490,"alternateName":491,"body":492,"description":558,"extension":84,"keywords":559,"meta":560,"navigation":87,"path":561,"seo":562,"stem":563,"updated":91,"__hash__":564},"glossary\u002Fglossary\u002Fen\u002Fmit-protocol-can.md","MIT Protocol (CAN Motor Control)","MIT 协议 \u002F MIT Mode",{"type":9,"value":493,"toc":554},[494,498,504,508,516,536,540],[12,495,497],{"id":496},"what-is-the-mit-protocol","What Is the MIT Protocol?",[17,499,500,501,503],{},"The ",[21,502,153],{}," (MIT Mode) is a CAN-bus motor control protocol originating from the MIT Mini Cheetah open-source legged robot. Its core idea: compress a full control command into one CAN frame carrying five quantities — target position p, target velocity v, position gain Kp, velocity gain Kd, and feed-forward torque τ.",[26,505,507],{"id":506},"hybrid-force-position-control","Hybrid Force-Position Control",[17,509,510,511,515],{},"The motor computes output torque as ",[512,513,514],"code",{},"τ_out = Kp·(p − p_actual) + Kd·(v − v_actual) + τ",". By tuning the gains, one protocol covers three control modes:",[38,517,518,524,530],{},[41,519,520,523],{},[21,521,522],{},"Pure torque control",": Kp = Kd = 0, feed-forward torque only — the basis of high-dynamic force control;",[41,525,526,529],{},[21,527,528],{},"Position control",": normal Kp\u002FKd values track position like a servo;",[41,531,532,535],{},[21,533,534],{},"Impedance control",": intermediate gains make the joint behave as a spring-damper for compliant contact.",[26,537,539],{"id":538},"why-it-became-a-de-facto-standard","Why It Became a De Facto Standard",[17,541,542,543,545,546,548,549,553],{},"The protocol is open, compact (one frame per command), and force-control-native. Many ",[54,544,111],{"href":67}," vendors now support it, creating a cross-brand control interface. All BXI ",[54,547,76],{"href":75}," support MIT-protocol CAN\u002FCANFD, and paired with the ",[54,550,552],{"href":551},"\u002Fen\u002Fmotors\u002Fcontrol-modules","PCIE-CAN control module"," achieve >1000 Hz whole-robot control loops.",{"title":79,"searchDepth":80,"depth":80,"links":555},[556,557],{"id":506,"depth":80,"text":507},{"id":538,"depth":80,"text":539},"The MIT protocol is a CAN-bus motor control scheme from the MIT Mini Cheetah open-source project, sending position, velocity, feed-forward torque, and gains in a single frame for hybrid force-position control — a de facto standard for robot joint motors.","MIT protocol, CAN motor control, hybrid force position control, motor communication protocol, CANFD",{},"\u002Fglossary\u002Fen\u002Fmit-protocol-can",{"title":490,"description":558},"glossary\u002Fen\u002Fmit-protocol-can","Q0AyuTGcRHiD67zY9FjjO_dfgKHkDH9mFB7PkB7eLds",{"id":566,"title":567,"alternateName":568,"body":569,"description":633,"extension":84,"keywords":634,"meta":635,"navigation":87,"path":636,"seo":637,"stem":638,"updated":91,"__hash__":639},"glossary\u002Fglossary\u002Fen\u002Fplanetary-gearbox.md","Planetary Gearbox","行星减速器",{"type":9,"value":570,"toc":629},[571,575,580,584,617,621,624],[12,572,574],{"id":573},"what-is-a-planetary-gearbox","What Is a Planetary Gearbox?",[17,576,19,577,579],{},[21,578,377],{}," is a gear train built from a central sun gear (input), several planet gears, and an outer ring gear. The planets spin on their own axes while orbiting the sun — like planets around a star, hence the name. It converts a motor's high speed into low-speed, high-torque output.",[26,581,583],{"id":582},"structural-advantages","Structural Advantages",[38,585,586,592,601,607],{},[41,587,588,591],{},[21,589,590],{},"Load sharing",": torque is split across multiple planet gears, giving high torque density and impact tolerance.",[41,593,594,597,598,600],{},[21,595,596],{},"Coaxial input\u002Foutput",": the compact cylindrical form factor suits integrated ",[54,599,68],{"href":67}," naturally.",[41,602,603,606],{},[21,604,605],{},"High efficiency",": single-stage efficiency typically exceeds 95%, higher than harmonic drives.",[41,608,609,612,613,58],{},[21,610,611],{},"Hollow-friendly",": a through-bore at the center enables ",[54,614,616],{"href":615},"\u002Fen\u002Fglossary\u002Fhollow-shaft-motor","hollow-shaft cable routing",[26,618,620],{"id":619},"planetary-vs-harmonic","Planetary vs. Harmonic",[17,622,623],{},"Harmonic drives offer large ratios (50–160) and zero backlash but lower efficiency, limited stiffness, and higher cost. Planetary gearboxes have smaller ratios (roughly 3–10 per stage, ~20 with compounding) and slight backlash, but win on efficiency, shock tolerance, and cost — decisive advantages for high-dynamic humanoid leg joints.",[17,625,72,626,628],{},[54,627,76],{"href":75}," all use a 19.5-ratio planetary design with a uniform 100 rpm rated output speed, so sizing reduces to picking the right torque tier.",{"title":79,"searchDepth":80,"depth":80,"links":630},[631,632],{"id":582,"depth":80,"text":583},{"id":619,"depth":80,"text":620},"A planetary gearbox uses a sun gear, planet gears, and a ring gear with multiple simultaneous tooth contacts and coaxial input\u002Foutput — delivering high torque density, compactness, and efficiency for robot joint motors.","planetary gearbox, planetary reducer, reduction ratio, robot gearbox, harmonic drive comparison",{},"\u002Fglossary\u002Fen\u002Fplanetary-gearbox",{"title":567,"description":633},"glossary\u002Fen\u002Fplanetary-gearbox","MMMfDWhs9jXejonqa7muAjm4xvqSAKEEFnKVsAI7_GE",{"id":641,"title":642,"alternateName":643,"body":644,"description":701,"extension":84,"keywords":702,"meta":703,"navigation":87,"path":704,"seo":705,"stem":706,"updated":91,"__hash__":707},"glossary\u002Fglossary\u002Fen\u002Fquasi-direct-drive.md","Quasi-Direct Drive (QDD) Actuator","准直驱执行器",{"type":9,"value":645,"toc":697},[646,650,660,664,670,681,685],[12,647,649],{"id":648},"what-is-a-quasi-direct-drive-actuator","What Is a Quasi-Direct Drive Actuator?",[17,651,19,652,655,656,659],{},[21,653,654],{},"quasi-direct drive (QDD) actuator"," pairs a large-diameter, high-torque motor with a ",[21,657,658],{},"low reduction ratio"," gearbox (typically 5–20), an approach popularized by legged-robot projects such as MIT Cheetah. \"Quasi-direct\" means close to direct drive (no reduction) while keeping a small amount of gearing.",[26,661,663],{"id":662},"the-key-property-backdrivability","The Key Property: Backdrivability",[17,665,666,667,36],{},"The higher the reduction ratio, the harder it is to drive the motor backwards from the output (reflected friction and inertia scale with the ratio squared). A low ratio keeps the joint ",[21,668,669],{},"backdrivable",[38,671,672,675,678],{},[41,673,674],{},"external impacts are absorbed by the motor yielding rather than the gears fighting them, protecting the drivetrain during landings and collisions;",[41,676,677],{},"joint forces can be estimated from motor current alone (proprioceptive force control), avoiding expensive torque sensors;",[41,679,680],{},"force-control bandwidth stays high — essential for jumping and running.",[26,682,684],{"id":683},"qdd-vs-high-ratio-drives","QDD vs. High-Ratio Drives",[17,686,687,688,690,691,694,695,451],{},"High-ratio drives (harmonic, cycloidal) win on torque density and holding efficiency but sacrifice backdrivability and impact tolerance; QDD is the opposite trade. Legged robots overwhelmingly choose QDD. The BXI ",[54,689,76],{"href":75}," use a 19.5-ratio ",[54,692,693],{"href":376},"planetary design"," — a classic QDD architecture with MIT-protocol force control (see ",[54,696,153],{"href":152},{"title":79,"searchDepth":80,"depth":80,"links":698},[699,700],{"id":662,"depth":80,"text":663},{"id":683,"depth":80,"text":684},"A quasi-direct drive actuator pairs a high-torque motor with a low reduction ratio (~5–20), balancing torque density with backdrivability and enabling force estimation from motor current — the dominant actuator scheme for dynamic legged robots.","quasi-direct drive, QDD, backdrivability, proprioceptive actuator, legged robot actuator",{},"\u002Fglossary\u002Fen\u002Fquasi-direct-drive",{"title":642,"description":701},"glossary\u002Fen\u002Fquasi-direct-drive","J8aqw1xT2DjByd9p_lLuB_Z_cRFCvjkWwND7pjhHDp8",{"id":709,"title":302,"alternateName":710,"body":711,"description":779,"extension":84,"keywords":780,"meta":781,"navigation":87,"path":782,"seo":783,"stem":784,"updated":91,"__hash__":785},"glossary\u002Fglossary\u002Fen\u002Fsim-to-real.md","仿真到实机迁移",{"type":9,"value":712,"toc":775},[713,717,722,726,733,756,760],[12,714,716],{"id":715},"what-is-sim-to-real","What Is Sim-to-Real?",[17,718,719,721],{},[21,720,302],{}," is the approach of training robot control policies in a physics simulator (such as MuJoCo or Isaac Gym\u002FLab) and then deploying them on real hardware. Training on real robots is slow, expensive, and breaks hardware; simulation runs thousands of robots in parallel at faster-than-real-time speed, which is what makes reinforcement learning of humanoid walking practical at all.",[26,723,725],{"id":724},"the-core-challenge-the-reality-gap","The Core Challenge: The Reality Gap",[17,727,728,729,732],{},"Simulation never matches reality exactly — friction, latency, motor characteristics, and sensor noise all differ. This mismatch is the ",[21,730,731],{},"sim-to-real gap",". The main mitigations:",[38,734,735,741,747],{},[41,736,737,740],{},[21,738,739],{},"Domain randomization",": randomly perturb simulation parameters (mass, friction, latency) during training, forcing the policy to become robust;",[41,742,743,746],{},[21,744,745],{},"System identification",": measure the real robot's motor response and inertial parameters accurately and write them back into the simulator;",[41,748,749,752,753,755],{},[21,750,751],{},"Actuator modeling",": model the torque-speed behavior of the ",[54,754,68],{"href":67}," explicitly — often the deciding factor for legged-robot transfer.",[26,757,759],{"id":758},"hardware-requirements","Hardware Requirements",[17,761,762,763,767,768,770,771,774],{},"Deployed policies command joints at hundreds of Hz, demanding high-bandwidth force control (see ",[54,764,766],{"href":765},"\u002Fen\u002Fglossary\u002Fquasi-direct-drive","quasi-direct drive"," and the ",[54,769,153],{"href":152},") and low-latency buses. The BXI ",[54,772,773],{"href":144},"Elf 3 humanoid"," ships with a MuJoCo environment and ROS2 SDK, and its >1000 Hz PCIE-CANFD control architecture supports the full simulation-to-hardware workflow.",{"title":79,"searchDepth":80,"depth":80,"links":776},[777,778],{"id":724,"depth":80,"text":725},{"id":758,"depth":80,"text":759},"Sim-to-Real is the approach of training robot control policies at scale in physics simulation and then transferring them to real robots — the dominant training paradigm for humanoid locomotion today.","sim-to-real, simulation to reality, reinforcement learning robotics, domain randomization, MuJoCo, Isaac",{},"\u002Fglossary\u002Fen\u002Fsim-to-real",{"title":302,"description":779},"glossary\u002Fen\u002Fsim-to-real","r1zVLzbVErZiDCC98F5liP8bIuQEPdAdUZGonwnDjak",{"id":787,"title":788,"alternateName":789,"body":790,"description":867,"extension":84,"keywords":868,"meta":869,"navigation":87,"path":870,"seo":871,"stem":872,"updated":91,"__hash__":873},"glossary\u002Fglossary\u002Fen\u002Fteleoperation.md","Teleoperation","遥操作",{"type":9,"value":791,"toc":863},[792,796,801,805,829,833,853],[12,793,795],{"id":794},"what-is-teleoperation","What Is Teleoperation?",[17,797,798,800],{},[21,799,788],{}," is the technique of a human operator controlling a robot's motion remotely in real time, with the robot executing tasks on site. The operator side (leader) captures human motion intent; the robot side (follower) reproduces it and streams back visual and other feedback, closing a human-in-the-loop control loop.",[26,802,804],{"id":803},"two-kinds-of-value","Two Kinds of Value",[806,807,808,814],"ol",{},[41,809,810,813],{},[21,811,812],{},"Immediate utility",": for tasks where autonomy isn't ready, teleoperation makes robots useful now — hazardous-environment work, remote assembly, live demonstrations.",[41,815,816,819,820,823,824,828],{},[21,817,818],{},"Data collection",": the observation–action trajectories produced by teleoperation are ",[21,821,822],{},"the highest-quality training data for imitation learning",". Today's mainstream ",[54,825,827],{"href":826},"\u002Fen\u002Fglossary\u002Fembodied-ai","embodied-AI"," training pipelines run on large-scale teleoperated demonstrations.",[26,830,832],{"id":831},"common-forms","Common Forms",[38,834,835,841,847],{},[41,836,837,840],{},[21,838,839],{},"Kinematically-matched leader arms",": a small leader arm mirroring the follower's structure — direct mapping, high precision;",[41,842,843,846],{},[21,844,845],{},"VR \u002F motion capture",": headsets and controllers, or full-body mocap, driving whole-body humanoid motion;",[41,848,849,852],{},[21,850,851],{},"Shared autonomy",": the human gives high-level commands while the robot handles balance and trajectories.",[17,854,72,855,859,860,862],{},[54,856,858],{"href":857},"\u002Fen\u002Frobots\u002Frobotic-arms","UpperBody 1 dual-arm platform"," ships with a plug-and-play operator console, and the ",[54,861,773],{"href":144}," supports both teleoperation and autonomous modes for manipulation data collection.",{"title":79,"searchDepth":80,"depth":80,"links":864},[865,866],{"id":803,"depth":80,"text":804},{"id":831,"depth":80,"text":832},"Teleoperation is real-time remote control of a robot by a human operator — both a practical deployment mode for humanoid robots today and the core method for collecting real-robot demonstration data to train embodied-AI models.","teleoperation, robot teleoperation, demonstration data collection, leader-follower control, imitation learning data",{},"\u002Fglossary\u002Fen\u002Fteleoperation",{"title":788,"description":867},"glossary\u002Fen\u002Fteleoperation","a0mBNQK0YBS6Xa9_ZlYjuzw1VfgVmXYcR5HIf1XT21A",1783425866674]