{
  "schemaVersion": 1,
  "id": "atlas-lesson-robot-observation-frame",
  "title": "Robot observations have coordinates and age",
  "category": "learning",
  "format": "Atlas lesson",
  "description": "Frame transforms, timestamps, measurement uncertainty and stale-state consequences.",
  "canonicalUrl": "https://semiaifoundry.com/learn/atlas/robot-observation-frame/",
  "publishedDate": null,
  "revisionDate": "2026-10-07",
  "author": null,
  "topic": "Observations become physical actions",
  "series": null,
  "license": null,
  "rightsUrl": "https://semiaifoundry.com/ai-atlas/licenses.html",
  "version": "1.3.0",
  "status": "Published educational lesson",
  "evidenceScope": "This is a worked explanation, not an executing external platform, model or robot. The quantities and fixtures are authored. Source scope is recorded in the automation learning review manifest; no source benchmark or private system behavior is claimed. Planar rigid transform, stated +90-degree convention, and constant speed bound over the age interval. No calibration, acceleration, sensor noise or live localization is solved.",
  "contentCoverage": "published-lesson",
  "artifactUrl": "https://semiaifoundry.com/ai-atlas/lesson-data/robot-observation-frame-40238bda81800030.json",
  "interactiveUrl": "https://semiaifoundry.com/ai-atlas/#robot-observation-frame",
  "related": [],
  "supersedes": [],
  "supersededBy": [],
  "contentHash": "a36a6b2e1698bfc9cde4668abc2665aed080e985a7353d36b632b3e64aa4f8a1",
  "contentLength": 6178,
  "jsonUrl": "https://semiaifoundry.com/agents/publications/atlas-lesson-robot-observation-frame.json",
  "markdownUrl": "https://semiaifoundry.com/agents/publications/atlas-lesson-robot-observation-frame.md",
  "sections": [
    {
      "id": "section-1",
      "title": "Robot observations have coordinates and age",
      "start": 53,
      "end": 461
    },
    {
      "id": "section-2",
      "title": "The mechanism",
      "start": 461,
      "end": 1681
    },
    {
      "id": "section-3",
      "title": "Science and assumptions",
      "start": 1681,
      "end": 1709
    },
    {
      "id": "section-4",
      "title": "Robot observations have coordinates and age · explicit relationship",
      "start": 1709,
      "end": 3062
    },
    {
      "id": "section-5",
      "title": "Engineering example",
      "start": 3062,
      "end": 3580
    },
    {
      "id": "section-6",
      "title": "Distinctions and tradeoffs",
      "start": 3580,
      "end": 3898
    },
    {
      "id": "section-7",
      "title": "Specifications and interfaces",
      "start": 3898,
      "end": 4363
    },
    {
      "id": "section-8",
      "title": "Interfaces",
      "start": 4363,
      "end": 4607
    },
    {
      "id": "section-9",
      "title": "Check your understanding",
      "start": 4607,
      "end": 4636
    },
    {
      "id": "section-10",
      "title": "Where is camera point (1,0) m after the stated +90-degree rotation and translation (2,3) m?",
      "start": 4636,
      "end": 4878
    },
    {
      "id": "section-11",
      "title": "Evidence and scope",
      "start": 4878,
      "end": 5466
    },
    {
      "id": "section-12",
      "title": "Sources",
      "start": 5466,
      "end": 5947
    },
    {
      "id": "section-13",
      "title": "Continue exploring",
      "start": 5947,
      "end": 5970
    },
    {
      "id": "section-14",
      "title": "Foundations",
      "start": 5970,
      "end": 6075
    },
    {
      "id": "section-15",
      "title": "Related lessons",
      "start": 6075,
      "end": 6178
    }
  ],
  "citations": [
    {
      "title": "Explore in Atlas",
      "url": "https://semiaifoundry.com/ai-atlas/#robot-observation-frame"
    },
    {
      "title": "Lynch and Park · Modern Robotics 3.3.1",
      "url": "https://modernrobotics.northwestern.edu/nu-gm-book-resource/3-3-1-homogeneous-transformation-matrices/"
    },
    {
      "title": "MoveIt Planning Scene",
      "url": "https://moveit.picknik.ai/main/doc/examples/planning_scene/planning_scene_tutorial.html"
    },
    {
      "title": "Published lesson data",
      "url": "https://semiaifoundry.com/ai-atlas/lesson-data/robot-observation-frame-40238bda81800030.json"
    },
    {
      "title": "Atlas releases and licensing",
      "url": "https://semiaifoundry.com/ai-atlas/licenses.html"
    },
    {
      "title": "Evaluation & the AGI frontier",
      "url": "https://semiaifoundry.com/learn/atlas/evaluation/"
    },
    {
      "title": "Tools, memory & agent loop",
      "url": "https://semiaifoundry.com/learn/atlas/tool-loop/"
    }
  ],
  "recordHash": "0a75d43c0961936cbe6ff8e985c2a1b264276b4c346c4507f60af56249365878",
  "contentFormat": "text/markdown",
  "content": "Observations become physical actions · Atlas v1.3.0\n\n# Robot observations have coordinates and age\n\nFrame transforms, timestamps, measurement uncertainty and stale-state consequences.\n\n[Explore in Atlas](https://semiaifoundry.com/ai-atlas/#robot-observation-frame)Logical state, task evidence and declared units; no physical assembly scale\n\nLearning objectiveTransform a planar observation with explicit frames and quantify a stated motion bound from its age.\n\n## The mechanism\n\nAn observation has a coordinate frame and a time as well as numeric values. A point measured by a camera cannot be used as a world position without the transform that relates those frames. Rotation changes orientation and translation changes origin; the order and frame convention matter. Measurements also become stale while the robot or environment moves. A controller using a correct coordinate transform from the wrong time can still act on an incorrect estimate. The worked planar transform is a small special case of the rigid-body transformations described in Modern Robotics. It is actual coordinate arithmetic, but it is not a camera calibration, localization system or live robot observation. Observation uncertainty and dynamic transforms require further model and evidence assumptions.\n\n1.  Each measurement names its source frame, units and acquisition time.\n2.  The declared rotation and translation map the point into the frame used for planning or control.\n3.  Observation age is computed from compatible timestamps rather than inferred from message arrival order.\n4.  Planning retains uncertainty and motion assumptions instead of treating a stale transformed point as current truth.\n\n## Science and assumptions\n\n### Robot observations have coordinates and age · explicit relationship\n\n$$\n\\mathbf p_W=R_{WC}\\mathbf p_C+\\mathbf t_{WC},\\qquad \\Delta d\\le v_{\\max}\\Delta t\n$$\n\nRotate a camera-frame point and add the camera origin in world coordinates; a stated speed bound limits motion over the observation age.\n\nSymbols and units\n\npW, pC\n\nThe same point expressed in world and camera coordinates\n\nm\n\nRWC\n\nRotation from camera axes to world axes\n\ndimensionless matrix\n\ntWC\n\nCamera origin expressed in world coordinates\n\nm\n\nvmax\n\nAssumed bound on relative speed\n\nm/s\n\nΔt\n\nObservation age in a compatible clock convention\n\ns\n\nΔd\n\nMotion-induced change bounded under the assumption\n\nm\n\nA point is at (1,0) m in a camera frame. The camera frame is rotated +90 degrees relative to the world and its origin is at (2,3) m in world coordinates. Multiplying by \\[\\[0,−1\\],\\[1,0\\]\\] and then adding (2,3) gives world point (2,4) m. Using (1,0) directly would place the point in the wrong frame. If the observation is 0.15 s old and relative speed is bounded by 0.2 m/s, motion alone can change separation by up to 0.03 m over that age; measurement and calibration error are additional.\n\nAssumptions Planar rigid transform, stated +90-degree convention, and constant speed bound over the age interval. No calibration, acceleration, sensor noise or live localization is solved.\n\n## Engineering example\n\nA point is at (1,0) m in a camera frame. The camera frame is rotated +90 degrees relative to the world and its origin is at (2,3) m in world coordinates. Multiplying by \\[\\[0,−1\\],\\[1,0\\]\\] and then adding (2,3) gives world point (2,4) m. Using (1,0) directly would place the point in the wrong frame. If the observation is 0.15 s old and relative speed is bounded by 0.2 m/s, motion alone can change separation by up to 0.03 m over that age; measurement and calibration error are additional.\n\n## Distinctions and tradeoffs\n\nCoordinate values have no unambiguous spatial meaning without their frame and time.\n\nMore accurate transforms reduce coordinate error but do not remove observation delay or unmodeled motion.\n\nRobotic AI must connect learned observations to the geometry and timing used by real actions.\n\n## Specifications and interfaces\n\nMechanism boundary\n\nWhere and when was the obstacle observed?\n\nPlanar rigid transform, stated +90-degree convention, and constant speed bound over the age interval. No calibration, acceleration, sensor noise or live localization is solved.\n\nEvidence type\n\nAuthored worked example\n\nWorked explanation with a bounded numeric or state example. This lesson does not claim an executing implementation of an external model or platform.\n\n### Interfaces\n\nDeclared input and state Input\n\nEach measurement names its source frame, units and acquisition time.\n\nChecked outcome Output\n\nTransform a planar observation with explicit frames and quantify a stated motion bound from its age.\n\n## Check your understanding\n\n### Where is camera point (1,0) m after the stated +90-degree rotation and translation (2,3) m?\n\n1.  (3,3) m\n2.  (2,4) m\n3.  (1,0) m\n\nAnswer and reasoning\n\n(2,4) m\n\nThe rotated point is (0,1); adding the world-frame translation gives (2,4).\n\n## Evidence and scope\n\nThis is a worked explanation, not an executing external platform, model or robot. The quantities and fixtures are authored. Source scope is recorded in the automation learning review manifest; no source benchmark or private system behavior is claimed. Planar rigid transform, stated +90-degree convention, and constant speed bound over the age interval. No calibration, acceleration, sensor noise or live localization is solved.\n\nThe worked example explains this mechanism. General numerical labs are a reference; they do not execute this lesson’s full mechanism.\n\n## Sources\n\n-   [Lynch and Park · Modern Robotics 3.3.1](https://modernrobotics.northwestern.edu/nu-gm-book-resource/3-3-1-homogeneous-transformation-matrices/)\n-   [MoveIt Planning Scene](https://moveit.picknik.ai/main/doc/examples/planning_scene/planning_scene_tutorial.html)\n\n[Published lesson data](https://semiaifoundry.com/ai-atlas/lesson-data/robot-observation-frame-40238bda81800030.json) · [Atlas releases and licensing](https://semiaifoundry.com/ai-atlas/licenses.html)\n\n## Continue exploring\n\n### Foundations\n\n-   [Evaluation & the AGI frontier](https://semiaifoundry.com/learn/atlas/evaluation/)\n\n### Related lessons\n\n-   [Tools, memory & agent loop](https://semiaifoundry.com/learn/atlas/tool-loop/)"
}