Skip to content
Marvin Bot

Sources

Every source cited in the guide (184 in all), grouped by kind, with what it was used for and the pages that cite it. Each playground also lists its own sources, numbered as they are cited on that page.

Papers

Section titled “Papers”
  1. A. A. Biewener, Biomechanical consequences of scaling, J. Exp. Biol. 2005 — how animals compensate for scaling · cited on 07
  2. A. Ghanbarzadeh, E. Najafi, Safe physical human-robot interaction through variable impedance control based on ISO/TS 15066, arXiv:2311.13814, 2023 — table of ISO/TS 15066 maximum permissible force, effective spring constant and effective mass per body region; vmax = F/√(μk); reduced mass. · cited on 12
  3. A. Kumar et al., RMA: Rapid Motor Adaptation for Legged Robots, RSS 2021 (arXiv 2107.04034) — online adaptation from history. · cited on 20
  4. A. L. Hof, M. G. J. Gazendam, W. E. Sinke, The condition for dynamic stability, J. Biomech. 2005 — extrapolated centre of mass · cited on 02
  5. Araujo et al., Retargeting Matters (GMR), arXiv:2510.02252 (2025) — GMR, non-uniform scaling, retargeting artefacts hurt tracking · cited on 17
  6. B. Katz, J. Di Carlo, S. Kim, “Mini Cheetah: A Platform for Pushing the Limits of Dynamic Quadruped Control,” ICRA 2019 — low-cost modular QDD actuators and the MIT-mode driver lineage. · cited on 04
  7. Ben et al., HOMIE, arXiv:2502.13013 — RL legs + exoskeleton arms, pose curriculum, height command; walk and squat with free arms · cited on 18, 22
  8. Black, Galliker, Levine, Real-Time Execution of Action Chunking Flow Policies, arXiv:2506.07339 — real-time chunking · cited on 18
  9. Christoph et al., ORCA hand, 2025 — tendon-driven open-source hand alternative · cited on 19
  10. D. Amodei et al., Concrete Problems in AI Safety, arXiv 1606.06565, 2016 — reward hacking as a general failure mode. · cited on 16
  11. D. Rempe, M. Petrovich, Y. Yuan et al., Kimodo: Scaling Controllable Human Motion Generation, arXiv 2603.15546, 2026 — Kimodo as kinematic motion diffusion · cited on 17
  12. F. Muratore et al., Robot Learning from Randomized Simulations: A Review, Frontiers in Robotics and AI 2022 (arXiv 2111.00956) — robustness versus conservatism of randomisation. · cited on 20
  13. G. Cervettini, R. Mauceri, A. Coppola, F. Bergonti, L. Fiorio, M. Maggiali, D. Pucci, “A Framework for Optimal Ankle Design of Humanoid Robots,” IEEE-RAS Humanoids 2025 (arXiv 2509.16469) — RSU and SPU parallel ankles, closed-form inverse kinematics, Jacobian rank loss, RSU judged better overall. · cited on 06
  14. Gu, Wang, Chen, Humanoid-Gym, arXiv 2404.05695 — sim-to-sim validation in MuJoCo; open humanoid RL framework with sim-to-sim check.; sim-to-sim validation in MuJoCo, humanoid randomisation practice. · cited on 14, 16, 20
  15. H. Wang, W. Zhang, R. Yu et al., PhysHSI: Towards a Real-World Generalizable and Natural Humanoid-Scene Interaction System, arXiv 2510.11072, 2025 — carry/sit skills as cited in roadmap · cited on 22
  16. He, Luo et al., OmniH2O, arXiv:2406.08858 — whole-body teleoperation tracker · cited on 18
  17. Huang et al., HoST: Learning humanoid standing-up control across diverse postures (2025) — get-up skill · cited on 22
  18. Hwangbo et al., Learning agile and dynamic motor skills for legged robots, 2019 — actuator and delay modelling for transfer; learned actuator networks; RL in simulation transferring to a legged robot.; actuator networks, actuator dynamics as a main gap.; measured actuator model used in training · cited on 03, 15, 16, 20, 22
  19. J. Denavit and R. S. Hartenberg, A kinematic notation for lower-pair mechanisms based on matrices, J. Applied Mechanics 1955 — classic joint-parameter convention · cited on 01
  20. J. Englsberger, C. Ott, A. Albu-Schäffer, Three-dimensional bipedal walking control based on DCM, T-RO 2015 — DCM dynamics · cited on 02
  21. J. Lee et al., Learning Quadrupedal Locomotion over Challenging Terrain, Science Robotics, 2020 (arXiv 2010.11251) — privileged teacher-student training. · cited on 16, 20
  22. J. Pratt, J. Carff, S. Drakunov, A. Goswami, Capture Point: a step toward humanoid push recovery, Humanoids 2006 — capture point, flywheel/hip strategy · cited on 02
  23. J. Schulman et al., High-Dimensional Continuous Control Using Generalized Advantage Estimation, arXiv 1506.02438, 2015 — GAE, the γ/λ trade-off. · cited on 16
  24. J. Schulman et al., Proximal Policy Optimization Algorithms, arXiv 1707.06347, 2017 — clipped surrogate objective, ε = 0.2, loss terms. · cited on 16
  25. J. Siekmann et al., Blind Bipedal Stair Traversal via Sim-to-Real Reinforcement Learning, RSS 2021 (arXiv 2105.08328) — biped stair policy transferred with randomisation. · cited on 20
  26. J. Tobin et al., Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World, IROS 2017 (arXiv 1703.06907) — domain randomisation idea.; origin of domain randomisation. · cited on 16, 20
  27. Kim et al., PyRoki, arXiv:2505.03728 (2025) — retargeting as cost minimisation, Levenberg-Marquardt in JAX · cited on 17
  28. L. Pinto et al., Asymmetric Actor Critic for Image-Based Robot Learning, arXiv 1710.06542, 2017 — critic with privileged observations. · cited on 16
  29. Liao et al., Berkeley Humanoid, 2024 (arXiv 2407.21781) — policy 50 Hz, PD 25 kHz on the motor driver · cited on 03
  30. Liao et al., BeyondMimic, 2025 (arXiv 2508.08241) — motion-tracking recipe; armature = I_rotor N², kp = I w², kd = 2 I zeta w, w = 10 Hz, zeta = 2, action scale 0.25 tau_max/kp, zero-armature effects; tracking formulation, observations, anchor, kernels, sigma values and weights, penalties, terminations, adaptive sampling · cited on 03, 15, 17
  31. Luo et al., SONIC, arXiv:2511.07820 — scale of a general tracker · cited on 18
  32. M. Mittal et al., Symmetry Considerations for Learning Task Symmetric Robot Policies, ICRA 2024 (arXiv 2403.04359) — mirror data augmentation. · cited on 16
  33. M. Vukobratović and B. Borovac, Zero-moment point: thirty five years of its life, IJHR 2004 — ZMP definition and balance condition · cited on 02
  34. Mahony, Hamel, Pflimlin, Nonlinear complementary filters on SO(3), IEEE TAC 2008 — complementary filtering · cited on 11
  35. N. Rudin et al., Learning to walk in minutes using massively parallel deep RL, CoRL 2021 — RL locomotion with random pushes; thousands of parallel robots on one GPU; policies output position targets to PD at a lower rate; massively parallel PPO, terrain curriculum.; friction, push and noise randomisation in GPU-parallel locomotion training.; GPU-parallel simulation makes sim-first cheap · cited on 02, 14, 15, 16, 20, 22
  36. NVIDIA, GR00T N1, 2025 (arXiv 2503.14734) — dual-system VLA design; System 1/2, Eagle-2 10 Hz, H=16, K=4, 63.9 ms L40, data pyramid · cited on 03, 18
  37. P. M. Wensing et al., Proprioceptive actuator design in the MIT Cheetah, T-RO 2017 — reflected inertia scales with N^2; low gear ratios; torque density and leg inertia · cited on 05, 07
  38. R. I. Davis, A. Burns, R. J. Bril, J. J. Lukkien, Controller Area Network (CAN) schedulability analysis: refuted, revisited and revised, Real-Time Systems 35(3), 2007 — worst-case frame length formula (g + 8s + 13 + ⌊(g + 8s − 1)/4⌋), response time growth with load. · cited on 10
  39. R. Riener, M. Rabuffetti, C. Frigo, Stair ascent and descent at different inclinations, Gait & Posture 2002 — human knee moments on stairs vs level walking · cited on 05
  40. R. Zanella, F. Califano, S. Stramigioli, Physical human-robot interaction: a critical review of safety constraints, arXiv:2601.19462, 2026 — transient limits taken as twice the quasi-static values except for the skull, forehead and face (no factor) · cited on 12
  41. Radosavovic et al., Real-World Humanoid Locomotion with RL, 2023 — further reading on learned humanoid controllers; humanoid RL locomotion, history-based policies.; humanoid sim-to-real with a history-conditioned policy. · cited on 03, 16, 20
  42. S. Kajita et al., Biped walking pattern generation by using preview control of ZMP, ICRA 2003 — cart-table model, preview control · cited on 02
  43. S. Kajita et al., The 3D linear inverted pendulum mode, IROS 2001 — LIPM · cited on 02
  44. S. Seok et al., “Design Principles for Energy-Efficient Legged Locomotion and Implementation on the MIT Cheetah Robot,” IEEE/ASME T-Mech, 2015 — large-gap motors with a single low-ratio stage.; motor torque scaling with gap radius · cited on 04, 07
  45. Sentis & Khatib, Synthesis of Whole-Body Behaviors through Hierarchical Control of Behavioral Primitives, IJHR 2(4) 2005 — prioritised whole-body control · cited on 18
  46. Shaw, Agarwal, Pathak, LEAP Hand, RSS 2023 — direct-servo open-source hand alternative · cited on 19
  47. T. He et al., ASAP: Aligning Simulation and Real-World Physics for Learning Agile Humanoid Whole-Body Skills, arXiv 2502.01143, 2025 — humanoid delta-action correction from real data. · cited on 20
  48. T. Koolen et al., Capturability-based analysis and control of legged locomotion, Part 1, IJRR 2012 — N-step capturability · cited on 02
  49. Tan et al., Sim-to-Real: Learning Agile Locomotion for Quadruped Robots, 2018 — latency as a sim-to-real factor; actuator identification, latency modelling, randomisation.; actuator identification to close the sim-to-real gap · cited on 03, 20, 22
  50. W. Zhao, J. P. Queralta, T. Westerlund, Sim-to-Real Transfer in Deep Reinforcement Learning for Robotics: a Survey, arXiv 2009.13303, 2020 — the reality gap and mitigation families. · cited on 20
  51. X. B. Peng et al., Sim-to-Real Transfer of Robotic Control with Dynamics Randomization, ICRA 2018 (arXiv 1710.06537) — dynamics randomisation.; dynamics randomisation with a memory policy. · cited on 16, 20
  52. Y. Chi et al., “Demonstrating Berkeley Humanoid Lite: An Open-source, Accessible, and Customizable 3D-printed Humanoid Robot,” RSS 2025 (arXiv 2504.17249) — cycloidal gearboxes chosen for printed parts; directly driven joints.; an open humanoid whose joints are each driven by their own actuator with no linkages. · cited on 04, 06
  53. Ze et al., TWIST, arXiv:2505.02833 — single whole-body controller, RL+BC · cited on 18
  54. Zhao et al., ACT, arXiv:2304.13705 — action chunking · cited on 18
  55. Zhu et al., Hiking in the Wild (2026) — depth-image locomotion · cited on 22

Software, models and official documentation

Section titled “Software, models and official documentation”
  1. build123d documentation — code-CAD alternative · cited on 08
  2. CadQuery documentation — code-CAD alternative · cited on 08
  3. candleLight_fw (gs_usb firmware) — the open gs_usb USB-CAN protocol supported by the Linux gs_usb driver. · cited on 10
  4. Isaac Lab docs, Teleoperation and Imitation Learning with Isaac Lab Mimic — demo generation · cited on 18
  5. Isaac Lab documentation — phase 1 simulation framework · cited on 22
  6. Isaac Lab documentation, Actuators — the DC-motor torque-speed clipping model.; DC motor torque-speed clipping, delayed actuators · cited on 04, 15
  7. Isaac Lab documentation, Importing a New Asset — URDF converter · cited on 13
  8. Isaac Lab documentation, isaaclab.envs.mdp (observation, reward and event terms) — term definitions, reward scaling by step time, base velocity task. · cited on 16
  9. Isaac Lab v3.0.0 release notes (2026-10-20) — Isaac Sim optional, physics backends; Newton and PhysX backends, MuJoCo Warp, Kamino experimental · cited on 13, 14
  10. Isaac Sim on GitHub — Isaac Sim source licence (Apache-2.0) and the Omniverse Kit SDK it downloads · cited on 14
  11. Linux Foundation, contribution of Newton — governance and co-developers · cited on 14
  12. Linux kernel, include/linux/can/length.h — classic and CAN FD field counts (41-bit FD extended header, CRC-17/21 fields with fixed stuff bits, 9-bit footer, 3-bit intermission). · cited on 10
  13. Manifold geometry library (manifold3d) — what manifold3d is · cited on 08
  14. mjlab, GitHub — open humanoid velocity task reference. · cited on 16
  15. MuJoCo documentation — cross-check physics engine · cited on 03
  16. MuJoCo documentation, Computation — soft contact, solvers, integrators, friction cones · cited on 14
  17. MuJoCo documentation, Modeling — MuJoCo loading URDF; solref/solimp meaning and defaults, timeconst >= 2 x timestep · cited on 13, 14
  18. MuJoCo documentation, XML Reference — armature, frictionloss, actuatorfrcrange, contype/conaffinity, position and motor actuators, compiler angle; armature · cited on 13, 15
  19. MuJoCo documentation, XML reference: equality constraints — closing kinematic loops with connect and weld constraints. · cited on 06
  20. mujoco-warp on PyPI (3.17.0, 2026-10-22) — MuJoCo Warp release status (development status 3 - Alpha), licence · cited on 14
  21. Newton physics engine, GitHub — licence, Warp, solvers, releases (v1.7.0, 2026-10-14) · cited on 14
  22. NVIDIA blog: Jetson T4000 and JetPack 7.1 — T4000 specs; modules on carrier boards · cited on 21
  23. NVIDIA Cosmos — Cosmos Transfer synthetic video · cited on 18
  24. NVIDIA GR00T-WholeBodyControl repository — decoupled WBC, SONIC checkpoints 50 Hz, look-ahead 80-200 ms · cited on 18
  25. NVIDIA Isaac GR00T — GR00T manipulation and synthetic data tools · cited on 22
  26. NVIDIA Isaac Sim documentation, Tutorial 10: Rig Closed-Loop Structures — articulations must be trees; one loop joint excluded from the articulation accumulates the most error. · cited on 06
  27. NVIDIA Isaac-GR00T repository and deployment benchmark — N1.7 architecture and latency table Benchmark table is in scripts/deployment/README.md; checkpoint nvidia/GR00T-N1.7-LIBERO libero_10 (32 DiT blocks). · cited on 18
  28. NVIDIA Kimodo repository — Kimodo capabilities, constraints, skeletons, licences · cited on 17
  29. NVIDIA PhysX 5 SDK documentation, Rigid Body Dynamics — PGS and TGS solvers · cited on 14
  30. NVIDIA PhysX SDK, GitHub — BSD-3-Clause, GPU source in 5.6 (April 2025) · cited on 14
  31. NVIDIA, Isaac GR00T N1.7 Hugging Face blog (17 Apr 2026) — 3B params, Cosmos-Reason2-2B, 20,854 h egocentric video, commercial licence · cited on 18
  32. NVIDIA, Isaac Lab documentation — parallel simulation for training · cited on 03
  33. NVIDIA, Isaac ROS release notes — 4.6 (Aug 2026) added Orin and JetPack 7.2 support; 5.0 moves to ROS 2 Lyrical · cited on 03
  34. NVIDIA, JetPack SDK — what JetPack is · cited on 03
  35. NVIDIA, Jetson AGX Orin Developer Kit User Guide — developer kit = module on a reference carrier · cited on 07
  36. NVIDIA, Jetson AGX Orin product page and technical brief — CPU, GPU, memory and 15-60 W power range · cited on 03
  37. NVIDIA, Jetson AGX Orin series technical brief — CPU/GPU/DLA specs, 275 TOPS split, 204.8 GB/s, 15-60 W · cited on 11
  38. NVIDIA, Jetson Linux Developer Guide: Controller Area Network (CAN) — two MTTCAN controllers on AGX Orin, classic up to 1 Mbit/s, FD data up to 15 Mbit/s, external 3.3 V transceiver, bring-up commands. · cited on 10
  39. NVIDIA, Jetson Linux Developer Guide: power modes for Jetson AGX Orin — AGX Orin 64 GB modes MAXN (no fixed budget), 15 W, 30 W (default), 50 W.; AGX Orin 64 GB power modes · cited on 09, 11
  40. NVIDIA, ProtoMotions — motion-tracking training framework; BeyondMimic-style experiment config, future frames, per-clip motion weights · cited on 03, 17
  41. Pollen Robotics, AmazingHand (CC-BY-4.0) — the hand meshes (simplified by the Marvin Bot project); 4 fingers, 8 DOF, two SCS0009 per finger parallel mechanism, 400 g, <200 EUR, licences, Enhanced STS3032 (<350 EUR), to-do; open hand design, build-from-parts option · cited on 00, 19, 21
  42. RobStride Product_Information — actuator models in the joint tree; CAN 2.0B 1 Mbit/s, 29-bit IDs, 14-bit encoder; peak, rated, no-load speed, ratio, mass and voltage per model; README-versus-manual differences.; RS04/RS03 peak, rated, speed, heat-sink condition; RS04 size, ratings, masses; torque constants, rated and peak torques, 60 V upper voltage limit (via description/actuators.yaml).; 1 Mbit/s classic CAN, 29-bit identifier layout, operation-control command and feedback reply (as implemented in robot/drivers/robstride.py).; 14-bit absolute encoder; peak torque and no-load speed behind effort and velocity limits; MIT operation mode, peak torque, speed, gear ratio; RS00/RS06/RS03 ratings; torque constant, CAN protocol and feedback frame.; actuator torque ratings · cited on 01, 03, 04, 05, 07, 09, 10, 11, 13, 15, 19, 20, 21
  43. RobStride, RS00 User Manual (CAN 2.0, rev. 260713) — RS00 Kt, currents, back-EMF, T-N curve. · cited on 04
  44. RobStride, RS03 User Manual (CAN 2.0, rev. 261013) — RS03 Kt, 17 V/krpm back-EMF, 200 rpm and 880 g, T-N curve, 215 × 220 mm plate, overload table. · cited on 04
  45. RobStride, RS04 User Manual (CAN 2.0, rev. 260713) — Kt 2.1 N·m/A, rated/peak currents, back-EMF, T-N curve, 345 / 220 mm plate ratings, 145 °C limit, overload and stall tables, operation-control (MIT) formula, field-weakening parameter. · cited on 04
  46. RobStride, RS06 User Manual (CAN 2.0, rev. 260713) — RS06 Kt, currents, back-EMF, T-N curve, test plate. · cited on 04
  47. ROS 2 Jazzy documentation — ROS 2 middleware · cited on 03
  48. ROS wiki, URDF joint and link elements — reading joint origins, axes and visuals; joint types, origin, axis, limit attributes and units · cited on 00, 13
  49. ROS wiki, urdf/XML/link — inertial, visual and collision elements · cited on 13
  50. ros2_control documentation (Jazzy) — controller manager, hardware interface concepts · cited on 03
  51. rsl_rl (Robotic Systems Lab, ETH Zurich), GitHub — PPO implementation, adaptive learning rate, symmetry augmentation. · cited on 16
  52. three.js r160 (MIT licence) — rendering the 3-D model in the browser · cited on 00

Datasheets, standards, books and other references

Section titled “Datasheets, standards, books and other references”
  1. 18650 Battery Store Molicel P45B — cell price · cited on 21
  2. Aerospace Metals, Aluminum 6061-T6/T651 datasheet — 6061-T6 yield 276 MPa, density 2.70, E 68.9 GPa · cited on 08
  3. AIFit Lab Damiao DM-J10010L-2EC — 120/40 N·m, ~$230 · cited on 21
  4. Analog Devices, LTC4359 ideal diode controller — what an ideal-diode OR does (a MOSFET acting as a low-loss diode). · cited on 09
  5. Asimov 1 documentation, Joint Design and Actuation — an open-source humanoid’s RSU ankle; its simulation models pitch and roll as direct joints mapped to two motors. · cited on 06
  6. Battery University, BU-304a Safety concerns with Li-ion — why packs need a BMS and current protection.; thermal runaway causes. · cited on 09, 12
  7. Battery University, BU-402 What is C-rate? — C-rate definition. · cited on 09
  8. Battery University, BU-702 How to store batteries — storage at about 40% state of charge. · cited on 12
  9. Battery University, BU-808 How to prolong lithium-based batteries — cycle life and available energy vs charge voltage (4.20 V 300-500 cycles; 4.13 V 400-700, 90%; 4.06 V 600-1,000, 81%).; lower charge voltage extends cycle life. · cited on 09, 12
  10. Birglen, Laliberte, Gosselin, Underactuated Robotic Hands, Springer 2008 — underactuation · cited on 19
  11. Boehm, A spiral model of software development and enhancement, IEEE Computer 1988 — risk-driven development · cited on 22
  12. Bosch Sensortec, BMI088 product page and flyer — gyro 0.014 dps/rtHz, accel 175 (flyer 230) ug/rtHz, ODR to 2 kHz, bias stability <2 deg/h · cited on 11
  13. C. Zhou, N. Tsagarakis, “On the Comprehensive Kinematics Analysis of a Humanoid Parallel Ankle Mechanism,” J. Mechanisms and Robotics 10(5), 2018 — forward kinematics needing a numeric solution; workspace and singularities of a 2-DOF parallel ankle. · cited on 06
  14. CAN in Automation, CAN knowledge — CAN arbitration and frames · cited on 03
  15. CAN in Automation, Designing a CAN network — CAN FD payload up to 64 bytes; data-phase rate limited by topology and transceivers. · cited on 10
  16. can-utils, canframelen.h (Czech Technical University in Prague) — confirmation that Linux bus-load tools use the same worst-case length (80 + 10s bits for 29-bit frames). · cited on 10
  17. CNX Software, NVIDIA increases Jetson prices, July 2026 — $3,499 dev kit price; AGX Orin dev kit $1,999 to $3,499; Orin 64 GB module $2,999; T4000 module $1,999 to $2,999 · cited on 11, 21
  18. D. A. Winter, Biomechanics and Motor Control of Human Movement, 4th ed., Wiley 2009 — segment heights as fractions of height (Fig. 4.1, after Drillis and Contini) · cited on 07
  19. Design World, How to calculate continuous and peak torque values for servo applications — RMS as equal-heating torque; peak and RMS checks · cited on 05
  20. Drake, model-based design and verification for robotics (MIT, project website) — hydroelastic contact, licence · cited on 14
  21. F. Hartwich, Bit time requirements for CAN FD, iCC 2013, CAN in Automation — bit-rate switching after arbitration. · cited on 10
  22. Feetech SCS0009 retailer specification (Maker Store) — 2.3 kg-cm stall at 6 V, ~0.75 rated, 11 g · cited on 19
  23. Feetech STS3032 retailer specification (Evelta) — 4.5 kg-cm stall, 1.5 rated at 6 V · cited on 19
  24. IEC 60204-1:2016 Safety of machinery: Electrical equipment of machines (catalogue page) — stop categories 0, 1 and 2. · cited on 12
  25. IEC 62133-2:2017 Safety requirements for portable sealed secondary lithium cells and batteries (catalogue page) — the battery safety standard to look for on bought cells. · cited on 12
  26. ISO 10218-1:2025 Robotics: Safety requirements, Part 1: Industrial robots (catalogue page) — scope of the industrial robot safety standard. · cited on 12
  27. ISO 13482:2014 Robots and robotic devices: Safety requirements for personal care robots (catalogue page) — existence and scope of a personal-care robot standard. · cited on 12
  28. ISO 13850:2015 Safety of machinery: Emergency stop function, principles for design (catalogue page) — what the e-stop standard covers. · cited on 12
  29. ISO/TS 15066:2016 Robots and robotic devices: Collaborative robots (catalogue page; standard text not reviewed) — body-region force thresholds, transient vs quasi-static contact, the speed-limit model. · cited on 12
  30. J. B. S. Haldane, On Being the Right Size, 1926 — square-cube argument · cited on 07
  31. K. M. Lynch and F. C. Park, Modern Robotics (2017, preprint) — DOF and Grübler’s formula, forward kinematics as a product of transforms, singularities; statics of open chains, tau = J^T F (section 5.2) · cited on 01, 05
  32. K. M. Lynch and F. C. Park, Modern Robotics: Mechanics, Planning, and Control, Cambridge University Press, 2017 — Jacobian-based inverse kinematics and static joint torques; torque, inertia, dynamics; forward kinematics, inertia, parallel-axis theorem; joint dynamics, reflected inertia; Jacobian, pseudo-inverse, IK; statics tau = J^T F, grasp friction · cited on 00, 07, 13, 15, 18, 19
  33. K. M. Lynch, F. C. Park, Modern Robotics, Cambridge University Press, 2017, §8.9 — gearing, reflected (apparent) inertia N²·J, harmonic drives.; velocity Jacobian, statics τ = JᵀF from virtual work, singularities, closed chains. · cited on 04, 06
  34. Katz, A low cost modular actuator for dynamic robots, MIT thesis 2018 — open-source QDD driver loops at tens of kHz · cited on 03
  35. Kollmorgen, Brushless servo motor sizing — RMS vs continuous rating, thermal time constant caveat · cited on 05
  36. Linux kernel documentation, SocketCAN — CAN as network interfaces, ip link configuration. · cited on 10
  37. Littelfuse, industrial and UL fuses catalogue (POWR-GARD JLLN Class T) — Class-T fuses’ DC rating (125-160 V DC) and 20 kA DC interrupting rating. · cited on 09
  38. Lynch and Park, Modern Robotics slides ch. 5 (statics) — Jacobian-transpose interpretation · cited on 05
  39. M. F. Ashby, Materials Selection in Mechanical Design, 4th ed. 2011 — strength and stiffness per mass · cited on 08
  40. Machinery Safety 101, Emergency stop series — plain-language definitions of the stop categories and the e-stop function. · cited on 12
  41. Mason, Starke, Komura, 100STYLE dataset (2022) — alternative motion source · cited on 17
  42. maxon, catalogue motor data page (thermal data block) — the convention of two thermal resistances and two time constants (winding, motor). · cited on 04
  43. MIT OCW 2.001 Mechanics and Materials I — beam bending, second moment of area, cantilever deflection · cited on 08
  44. MIT OCW 2.004 Dynamics and Control II — second-order systems, overshoot formula · cited on 15
  45. Molicel, INR-21700-P45B product page — 3.6 V nominal, 4.5 Ah, 16.2 Wh, 45 A max continuous, 242 Wh/kg, 13.8 mΩ max internal resistance. · cited on 09
  46. NASA ESTO Technology Readiness Levels — component to system test ladder · cited on 22
  47. NASA Work Breakdown Structure Handbook — cost estimates hang from a product-oriented breakdown · cited on 21
  48. NBS Handbook 100, Copper Wire Tables (1966) — copper’s temperature coefficient of resistance, 0.00393 per K at 20 °C. · cited on 04
  49. ONNX Runtime documentation — running ONNX policies · cited on 03
  50. OpenAI Spinning Up, Proximal Policy Optimization — on-policy policy gradient explanation. · cited on 16
  51. OpenELAB, RobStride04 complete guide (reseller specification) — RS04 terminal resistance 0.16 Ω ±10% (reseller figure, not from RobStride’s manual). · cited on 09
  52. OpenSCAD, the programmers’ solid 3D CAD modeller (project website) — code-CAD alternative · cited on 08
  53. OpenUSD, Introduction to USD — what USD is · cited on 13
  54. Orbbec Gemini 335L product page — camera 124 x 29 x 27 mm, 133 g · cited on 07
  55. Orbbec store Gemini 335L — $359 list · cited on 21
  56. Orbbec, Gemini 330 series datasheet V1.6 — 335L baseline 95 mm, FOV, precision, ranges, min-Z, MX6800, power, size, mass · cited on 11
  57. P. M. Wensing, A. Wang, S. Seok, D. Otten, J. Lang, S. Kim, “Proprioceptive Actuator Design in the MIT Cheetah: Impact Mitigation and High-Bandwidth Physical Interaction for Dynamic Legged Robots,” IEEE T-RO 33(3), 2017 — proprioceptive actuation, reflected inertia and impact mitigation, backdrivability.; keeping leg inertia low by placing motors near the body and driving distal joints through linkages. · cited on 04, 06
  58. Pilkey, Pilkey, Bi, Peterson’s Stress Concentration Factors, 4th ed. 2020 — Kt of 2-3 at holes and notches · cited on 08
  59. Polymaker PolyMide PA6-CF TDS V4.1 — PA6-CF tensile XY 82 / Z 64 MPa (conditioned), E 5.7/4.7 GPa, density 1.17 · cited on 08
  60. Prusament PETG technical data sheet — PETG printed yield 47-50 MPa, density 1.27 · cited on 08
  61. Quickparts, Aluminum 7075 datasheet — 7075-T6 yield 503 MPa, density 2.81, E 71.7 GPa · cited on 08
  62. R. Drillis, R. Contini, M. Bluestein, Body segment parameters: a survey of measurement techniques, Artificial Limbs 1964 — origin of the segment fractions · cited on 07
  63. R. Kumar et al., “Kinematic analysis of a novel parallel 2SPRR+1U ankle mechanism in humanoid robot,” Advances in Robot Kinematics 2018 — parallel ankle kinematics, workspace and singularity analysis. · cited on 06
  64. R. S. Sutton, A. G. Barto, Reinforcement Learning: An Introduction, 2nd ed., MIT Press, 2018 — RL vocabulary, return and discount, policy gradient. · cited on 16
  65. R. Tedrake, Underactuated Robotics (MIT 6.832), chapter “System Identification” — least-squares identification of models linear in parameters. · cited on 20
  66. R. Tedrake, Underactuated Robotics, ‘Humanoid robots’ — floating base, underactuation; CoP/ZMP on flat ground, support polygon, pendulum equation · cited on 01, 02
  67. Robert Bosch GmbH, CAN Specification version 2.0 (1991) — frame fields, standard and extended identifiers, arbitration, bit stuffing rule. · cited on 10
  68. RobotShop CubeMars AK80-64 — 120/48 N·m, ~$890-1,240 · cited on 21
  69. S.C. Code Regs. § 8-1210, IRC R311.7.5.1 Risers, as adopted in South Carolina (Cornell LII) — 196 mm maximum riser, 254 mm minimum tread · cited on 05
  70. Seeed Studio RobStride 03 listing — $225, $214 at 10+, 880 g · cited on 21
  71. Seeed Studio RobStride 04 listing — $255, $242 at 10+ · cited on 21
  72. Seeed Studio wiki, GR00T N1.7 TensorRT on AGX Orin (JetPack 7.2) — 0.2755 s per 16-action chunk on AGX Orin; 0.2755 s per 16-action chunk · cited on 03, 11
  73. SparkFun Jetson AGX Orin 64GB Developer Kit — $3,499 listing · cited on 21
  74. Szeliski, Computer Vision: Algorithms and Applications, 2nd ed. — depth from disparity and its error · cited on 11
  75. Texas Instruments, Introduction to the Controller Area Network (CAN), SLOA101B — differential signalling, dominant/recessive levels, 120 Ω termination. · cited on 10
  76. Texas Instruments, TIDUF56 design guide (field weakening and MTPA) — field weakening extends operation above base speed. · cited on 04
  77. USITC Harmonized Tariff Schedule — duty depends on classification and country (slider hypothetical) · cited on 21
  78. RobStride, RS10P User Manual (CAN 2.0, rev. 260907) and RS series specification sheet (2026-10-09) — RS10P torque, speed, ratio, mass, Kt, winding resistance, overload tables, output-side inertia, MIT ranges. · cited on 04, 09