{
  "title": "Otterchip Model E — provisional visual and pose review (phase 2)",
  "date": "2026-11-06",
  "reviewer": "Claude Opus 5.5 (claude-opus-5-5), independent reviewer; writes only hardware/otterchip_model_e/review/**",
  "status": "PROVISIONAL. The optimizer is still running. The E dimensions reviewed here are neither final nor shown to be globally optimal. Mechanical CAD is not ready, so the production review is pending.",
  "designReviewed": {
    "source": "native-browser-cases.json (16 cases, M/E × ages 3,6,9,12 × typing/music, browser errors = [])",
    "eParamsChangedFromM": { "height": "250 (M 221)", "tilt": "20° (M 10.5°)", "deck": "62.5 flat (M 91, slope 4)", "mididrop": "23 (M 20)", "padrise": "7 (M 11)", "padgap": "10 (M 14.5)", "padcrown": "1 (M 1.5)", "rear": "100 (M 92)", "blend": "70 (M 67)", "grillecorner": "4 (M 0)", "controledge": "80 (M 47)", "controlheight": "30 (M 17.5)" },
    "unchanged": "width 355, setback 159, 10.5\" display, underlap 39, GK64X and LPK25 component sizes",
    "overallMm": { "E": [355, 342.0, 405.4], "M": [355, 316.9, 382.2] },
    "portPlan": "Headphone pair moves to the left side and the power switch to the right side (coordinator proposal)."
  },
  "imagesRead": [
    "/tmp/otterchip-e-review-E-body.png",
    "/tmp/otterchip-e-review-M-body.png",
    "/tmp/otterchip-e-review-E-3-typing-side.png",
    "/tmp/otterchip-e-review-E-3-music-side.png",
    "/tmp/otterchip-e-review-E-6-typing-threequarter.png",
    "/tmp/otterchip-e-review-E-6-music-threequarter.png",
    "/tmp/otterchip-e-review-E-9-typing-side.png",
    "/tmp/otterchip-e-review-E-12-typing-side.png",
    "/tmp/otterchip-e-review-E-12-music-side.png",
    "/tmp/otterchip-e-review-M-6-typing-side.png"
  ],
  "imagesNotRead": "The other M/E age and activity captures were not opened. Their raw metrics were read from native-browser-cases.json.",
  "rawMetrics": {
    "note": "Values come from the app's own solver output in native-browser-cases.json, using auto furniture with the same algorithm for both models; furniture heights are nearly identical for M and E. Incidence is my recomputation: the angle between (eye − screen centre) and the exported screen normal. Wrist reach is my unitless ratio of solved right shoulder→wrist distance to (upper arm + forearm + hand). Fingertip reach uses the same denominator. These are geometric reference values, not clinical outcomes.",
    "columns": ["model", "age", "activity", "tableMm", "viewMm", "gazeDeg", "neckDeg", "incidenceDeg", "screenTopBelowEyeMm", "elbowDeg", "wristDeg(+ext)", "forearmDeg(+rising)", "shoulderFlexDeg", "keyAboveNeutralMm", "reachErrMm", "penetrationMm", "shoulderToWrist/armChain", "shoulderToFingertip/armChain", "status"],
    "rows": [
      ["M", 3, "typing", 344, 379.6, 14.2, 0.0, 3.7, 12.7, 93.9, -8.5, 27.7, 31.6, 112, 0, 0.3, 0.538, 0.754, "conflict"],
      ["E", 3, "typing", 344, 364.6, 10.2, 0.0, 9.8, -12.7, 95.5, -2.0, 19.5, 25.0, 83, 0, 0.0, 0.545, 0.756, "conflict"],
      ["M", 3, "music", 340, 437.0, 12.9, 0.0, 2.4, 16.7, 89.2, 1.0, 11.9, 11.0, 54, 0, 0.0, 0.517, 0.726, "conflict"],
      ["E", 3, "music", 341, 417.3, 9.3, 0.0, 10.7, -9.7, 88.9, 1.0, 6.6, 5.5, 30, 0, 0.0, 0.516, 0.724, "caution"],
      ["M", 6, "typing", 421, 460.7, 20.3, 0.3, 9.8, 78.9, 98.3, -3.0, 20.3, 28.6, 111, 0, 0.0, 0.562, 0.772, "conflict"],
      ["E", 6, "typing", 423, 419.9, 17.9, 0.0, 2.1, 51.9, 90.6, 1.0, 15.1, 15.7, 84, 0, 0.0, 0.529, 0.732, "conflict"],
      ["M", 6, "music", 419, 500.7, 18.9, 0.0, 8.4, 81.3, 88.9, 1.0, 9.7, 8.5, 53, 0, 0.0, 0.521, 0.723, "conflict"],
      ["E", 6, "music", 420, 478.2, 16.0, 0.0, 4.0, 54.9, 88.8, 0.5, 5.5, 4.3, 29, 0, 0.0, 0.521, 0.724, "caution"],
      ["M", 9, "typing", 493, 486.3, 24.8, 4.8, 14.3, 123.6, 89.0, 1.0, 18.9, 17.9, 120, 0, 0.0, 0.527, 0.722, "conflict"],
      ["E", 9, "typing", 493, 468.8, 22.5, 2.5, 2.5, 102.1, 89.0, 1.0, 14.7, 13.7, 91, 0, 0.0, 0.527, 0.722, "conflict"],
      ["M", 9, "music", 489, 555.5, 22.5, 2.5, 12.0, 131.6, 89.2, 1.0, 9.2, 8.3, 59, 0, 0.0, 0.527, 0.724, "conflict"],
      ["E", 9, "music", 490, 538.3, 20.2, 0.2, 0.2, 108.7, 89.2, 0.5, 5.6, 4.8, 35, 0, 0.0, 0.527, 0.725, "caution"],
      ["M", 12, "typing", 555, 533.2, 28.8, 8.8, 18.3, 176.4, 89.0, 1.0, 16.7, 15.7, 118, 0, 0.0, 0.529, 0.721, "conflict"],
      ["E", 12, "typing", 556, 513.6, 26.7, 6.7, 6.7, 154.0, 89.1, 1.0, 12.9, 12.0, 90, 0, 0.0, 0.530, 0.722, "conflict"],
      ["M", 12, "music", 552, 601.0, 26.2, 6.2, 15.7, 184.4, 89.0, 0.5, 8.1, 7.2, 57, 0, 0.0, 0.530, 0.724, "conflict"],
      ["E", 12, "music", 553, 579.9, 24.3, 4.3, 4.3, 161.4, 89.0, 0.5, 4.6, 3.6, 32, 0, 0.0, 0.529, 0.724, "caution"]
    ]
  },
  "achievedPoseChanges": {
    "framing": "These are raw geometric changes, M → provisional E. A lower composite score or fewer 'conflict' flags is not evidence of clinical or comfort benefit.",
    "keysAboveNeutral": "Typing improves by 28–29 mm at every age (112→83, 111→84, 120→91, 118→90). Music improves by 22–24 mm (54→30, 53→29, 59→35, 57→32). Typing is still flagged 'Keys sit high for relaxed arms' at all ages.",
    "forearmRise": "Typing falls by 3.8–8.2° (age 3: 27.7→19.5; age 12: 16.7→12.9) but still rises 13–20° toward the keys. Music falls by 3.5–5.3° to +4.6…+6.6°.",
    "shoulderFlex": "Typing: age 3 31.6→25.0, age 6 28.6→15.7, age 9 17.9→13.7, age 12 15.7→12.0. Music: 3.6–5.5° in E.",
    "wristAndElbow": "At age 3 typing, wrist flexion goes −8.5→−2.0° and elbow 93.9→95.5°. Elsewhere the wrist is ≈ +1° and the elbow ≈ 89–91° in both models.",
    "viewing": "Gaze drops 2–4° lower in E (age 12 typing 28.8→26.7°). Head flex at age 12 typing goes 8.8→6.7°. Incidence improves at ages 6–12 (typing 9.8–18.3° → 2.1–6.7°) but worsens at age 3 (3.7→9.8° typing, 2.4→10.7° music).",
    "tradeoffs": "1) Viewing distance is SHORTER in E by 15–41 mm (typing age 6: 461→420 mm) because the arm-preferred seat moves closer. The adult 50–100 cm band is reference-only for children. 2) Screen top is ABOVE eye level for the age-3 reference child in E (12.7 mm typing, 9.7 mm music; M was 12.7/16.7 mm below). This sits in the solver's caution band, against the OSHA adult top-at-or-below-eye reference. It comes from the taller 250 mm screen height. 3) At age 3 the wrist-to-pad gap is 0 mm in E typing (M −0.3 mm), meaning the wrist is at pad contact in the static pose. That is consistent with resting on the pad during pauses, and also means a smaller hover margin.",
    "reachNormalisation": "Shoulder→wrist / arm chain is 0.52–0.56 and shoulder→fingertip / arm chain is 0.72–0.77 in both models. E changes it by at most −0.04 (age 6 typing 0.772→0.732). Hands stay in a mid-reach zone; no overreach is indicated for home/representative targets. Far-key and control reach remain untested secondary checks."
  },
  "visualObservations": [
    { "id": "O1", "image": "E-body vs M-body", "observation": "E reads as one fluid fixed body: a lower keyboard deck, a taller screen with steeper (20°) tilt, and a deeper concave sweep between deck and display. The front of the music bay is now clean, with the front jack and switch plates removed, which is a clear visual improvement over M's front port plates." },
    { "id": "O2", "image": "E-body", "observation": "The cream blank sweep between the GK64X top row and the display's bottom bezel is noticeably larger than on M (roughly 2× in the three-quarter view). With the taller rear, it gives E a heavier 'tombstone' upper mass over a low deck. Balanced but monumental; the large uninterrupted surface may read as empty." },
    { "id": "O3", "image": "E-body (also M-body)", "observation": "In the three-quarter view, the speaker grille panels do not appear to share top and bottom lines with the display bezel. The left grille seems to end above the screen bottom, and the right grille extends slightly below it. This is present in M too, and may be a perspective effect of grilles on cheek surfaces versus a 20°-tilted screen. Unconfirmed; check in an orthographic front view." },
    { "id": "O4", "image": "E side views (3/9/12)", "observation": "The side slab is a large flat taupe face with three features: a small rectangular feature at screen height, and near the base a small dark plate plus a short dash. The root is already removing the generic legacy USB rectangles, so that is not raised here. The upper rectangle is present on M as well, and its identity cannot be confirmed from images. The proposed left-side headphone pair and right-side power switch are small at this scale and sit low near the table." },
    { "id": "O5", "image": "E-6-typing-threequarter", "observation": "The palm pad reads as a dark, slim, flush element in front of the typing keys over the music cover. The lower 7 mm rise gives a calmer, more integrated front than M. Hands hover above the keys with no visible intersection." },
    { "id": "O6", "image": "E-6-music-threequarter, E-3/12-music-side", "observation": "With the cover off, hands sit over the exposed LPK25 white keys in front of the typing bridge. Fingers appear visually near the bridge edge in the side views. Exported penetration is 0 mm, and 2D overlap is not treated as collision. The bridge fingertip-capsule check (initial review V6) is still not run." },
    { "id": "O7", "image": "all person images", "observation": "No real intersections are visible in the three-quarter views. The 2D overlap of hands and keys in side views matches exported reach error 0 mm and penetration 0 mm in all 8 E cases. Status chips: typing shows 'Keys sit high for relaxed arms' at all ages; E music shows 'Outside a reference posture'." }
  ],
  "prioritizedIssues": [
    { "priority": "P1", "id": "I1", "title": "Age-3 screen top above eye level, and worse incidence at age 3", "type": "ergonomic tradeoff (raw)", "detail": "Comes from height 250 / tilt 20. The optimizer objective should see this as part of the grouped viewing term, not as a hard child limit. It should be stated plainly in comparisons." },
    { "priority": "P1", "id": "I2", "title": "Bridge fingertip-capsule check still pending for music", "type": "geometric verification", "detail": "Required before claiming music fit at underlap 39." },
    { "priority": "P2", "id": "I3", "title": "Large blank sweep and tall rear mass", "type": "visual balance", "detail": "See recommendations R1–R3. Changes that keep the geometry need no re-optimisation." },
    { "priority": "P2", "id": "I4", "title": "Grille/display line alignment", "type": "visual detail, unconfirmed", "detail": "Verify orthographically, then align grille top and bottom to the display bezel or active area in its tilted plane." },
    { "priority": "P3", "id": "I5", "title": "Side controls: identity, legibility and plug clearance", "type": "practical/visual", "detail": "Confirm the upper side rectangle's purpose. Group the left headphone pair on one recessed plate with an icon. Mechanical review must still confirm room for the plug body and cable bend above the table." },
    { "priority": "P3", "id": "I6", "title": "Shorter viewing distance in E", "type": "ergonomic tradeoff (raw)", "detail": "15–41 mm closer than M. Report it raw with the adult band shown only as a reference." }
  ],
  "blockingVisualError": { "value": false, "basis": "No real intersection is visible in the three-quarter views, and exported reach error and penetration are 0 for all E cases. The known legacy USB rectangles are excluded as already being fixed. Items I1/I2 are ergonomic or verification items, not rendering errors." },
  "recommendations": {
    "geometryPreserving": [
      "R1 Sweep surface: add a subtle tonal split, or a slightly darker satin band, along the concave sweep, following the blend curve so the surface reads as intentional. Alternatively, add a fine engraved logotype or pin-light status line at about 40% of sweep height. Avoid ribs or grooves where children's palms or fingers may slide.",
      "R2 Edge treatment: soften the top-rear and side-cheek corner radii consistently (match the 4 mm grillecorner and pad corner family). Use a continuous radius along the sweep-to-cheek intersection so the side silhouette reads as one bent sheet.",
      "R3 Finish: matte cream body with a slightly lower-gloss taupe side panel. Keep the dark grille/pad accents in one shared tone so pad, grilles and display bezel read as a set.",
      "R4 Grille coordination: once alignment is verified, snap grille top and bottom to the display bezel lines and share the hole pitch with any side vents. If cosmetic only, keep the acoustic open area unchanged.",
      "R5 Side ports: put the headphone pair on a single recessed, labelled plate and the power rocker in a matching recessed plate on the right. Keep the small upper rectangle only if functional.",
      "R6 Music cover and pad: keep the pad flush with a 7 mm rise. Give the music-cover pull tabs the same corner family and accent colour as the pad edge."
    ],
    "requireObjectiveReevaluation": [
      "Any change to display height, tilt, setback, deck height or slope, typing drop, underlap, pad rise or gap, or front lip depth. These move eye, screen, key or obstacle geometry.",
      "Lowering the screen to reduce the tall mass or the age-3 above-eye top: re-run the viewing group and incidence at all ages and statures.",
      "Extending grilles or cheeks forward or downward, if they enter the arm or hand sweep near the keys or change the front lip.",
      "Any rib, lip or feature added on the deck or pad within the forearm/hand corridor, because it changes the collision checks."
    ]
  },
  "chartPresentationRequirements": [
    "Plot raw metrics per age (x = 3..12) for M and E typing, and M and E music. D and W appear typing-only, with music shown as n/a rather than zero.",
    "Show the composite last, with its weights stated and a weight-sensitivity band. Label it as a designer heuristic, not a health score.",
    "Show adult OSHA bands (viewing 50–100 cm, gaze 15–20°, top at or below eye level) as reference lines labelled 'adult reference'.",
    "Include incidence and screen-top-vs-eye together with gaze, grouped as viewing.",
    "Show wrist reach as a unitless ratio (shoulder→wrist / arm chain), not raw mm.",
    "Mark stature sensitivity (P5/P95 or ±10%) as a band if available, and do not call it percentiles of body proportion.",
    "Unfinished chart UI is not judged here."
  ],
  "knownLimitations": [
    "The E dimensions are provisional and the optimizer is still running.",
    "Mechanical CAD is not ready, so the production, assembly and serviceability review is pending.",
    "Only 10 images were viewed. Visual judgments are from rendered previews under one lighting and palette.",
    "The person model is an existing approximation. Static single poses do not model typing or playing movement.",
    "Incidence and reach ratios are my recomputations from exported points and the screen normal.",
    "The grille misalignment and side-feature identity are unconfirmed from perspective views."
  ]
}
