Robot Wiki

Credits

Robot Wiki uses real photographs and diagrams, and every one of them is listed here with its creator, the page it came from, and the licence that permits its reuse. Article images stay under CC0, CC BY, CC BY-SA, public domain, or a documented reuse permission. Market-map logos also include official company marks recorded as unlicensed or unknown when no reuse grant is named. No image is AI-generated. The site's own text and original diagrams are available under CC BY 4.0.

Written and maintained by Josef Chen (@josefchen).

Who is behind this wiki

I am Josef Chen. Spent 3 years building and deploying robots at KAIKAKU (acquired by REEF). What I needed over those years was scattered across papers that skip the deployment context and demos that never quote a failure rate, so I kept rebuilding the same mental map of the field from scratch. This wiki is that map, written down for engineers arriving from machine learning. Every claim links to the source it came from, so you can check it yourself.

Corrections and source disputes are welcome. Read the editorial policy or report a factual issue on GitHub.

  1. A yellow Boston Dynamics Spot quadruped robot walks across an airfield tarmac alongside Royal Air Force airmen in camouflage uniform.
    A Boston Dynamics Spot quadruped walks with Royal Air Force airmen at RAF Leeming during the Agile Liberty 21-2 technology trials in August 2021.
    Photo: Senior Airman John Ennis, U.S. Air Force / Wikimedia Commons. Licence: Public domain.

    Appears on: Home

  2. The 2013 DARPA Atlas humanoid robot stands facing forward with its exposed metal frame, hydraulic lines, and stereo camera head visible.
    The original Atlas humanoid, built for the 2013 DARPA Robotics Challenge: hydraulically actuated, tethered, with its frame and sensor head exposed. The electric Atlas discussed above is its clean-sheet successor.
    Photo: DARPA / Wikimedia Commons. Licence: Public domain.

    Appears on: Hardware Taxonomy

  3. A stroboscopic NASA photograph shows a PUMA 560 robot arm in several successive positions as it moves small parts between fixtures on a table.
    A PUMA 560 at NASA Ames, photographed in a stroboscopic exposure as it cycles through its workspace. The six-revolute PUMA layout is the canonical teaching example for the Denavit-Hartenberg convention.
    Photo: NASA / Dominic Hart / Wikimedia Commons. Licence: Public domain.

    Appears on: Kinematics

  4. A red and grey ANYmal quadruped robot stands on a metal grate walkway inside an industrial facility, surrounded by pipes and machinery.
    ANYmal, the ANYbotics quadruped behind the ETH Zurich locomotion lineage, on an industrial inspection floor. The actuator-network and terrain-curriculum results above all ran on this platform.
    Photo: ANYbotics / Wikimedia Commons. Licence: CC BY-SA 4.0.

    Appears on: Legged Locomotion Lineage

  5. A white seven-axis Franka Emika Panda robot arm mounted on a demonstration table at CeBIT 2017, guided by hand above trays of small parts.
    The Franka Emika Panda, the seven-DoF research arm, demonstrated at CeBIT 2017 with a human guiding it by hand.
    Photo: Ims / Wikimedia Commons. Licence: CC BY-SA 4.0.

    Appears on: Hardware Taxonomy

  6. Original schematicDiagram of covariate shift: demonstration trajectories form a narrow corridor around the expert path, while the policy rollout starts inside the corridor and drifts outside it.
    Demonstrations cover a narrow corridor of states around the expert path. The policy's first mistake takes it outside that corridor, where its next mistake is more likely. Deviation feeds itself.
    Diagram: Robot Wiki contributors / Robot Wiki (original diagram). Licence: CC BY 4.0.

    Appears on: Behavior Cloning Foundations

  7. Original schematicDiagram of temporal ensembling: three overlapping action chunks each contain a prediction for the same action a_t, and oldest-to-newest unnormalized weights are 1.00, 0.61, and 0.37 for illustrative m=0.5; the reference code uses m=0.01.
    Three chunks issued at t-2, t-1, and t predict the current action. Oldest-to-newest raw weights are 1.00, 0.61, and 0.37 for illustrative m=0.5; divide by their sum before averaging. This emphasizes the reference convention: oldest gets the largest weight. The pinned ACT code uses m=0.01, not 0.5.
    Diagram: Robot Wiki contributors / Robot Wiki (original diagram). Licence: CC BY 4.0.

    Appears on: Action Chunking (ACT and ALOHA)