{
  "source": "Cooked Index — occupational AI risk register",
  "page": "https://cookedindex.com/jobs/textile-knitting-and-weaving-machine-setters-operators-and-tenders/",
  "methodology": "https://cookedindex.com/methodology",
  "notice": "Verdicts are re-examined as evidence accumulates. Re-fetch before relying on this; the page above always carries the current score.",
  "scored_at": "2026-08-11",
  "model": "claude-opus-5",
  "occupation": {
    "title": "Textile Knitting and Weaving Machine Setters, Operators, and Tenders",
    "soc_code": "51-6063",
    "category": "Production",
    "us_employment": 13030,
    "median_annual_wage": 39530
  },
  "verdict": "COOKED",
  "risk_resistance": 31,
  "contested": false,
  "near_boundary": false,
  "dimensions": {
    "task_resistance": 10,
    "embodiment": 13,
    "liability_shield": 1,
    "trust_premium": 2,
    "judgment_accountability": 5
  },
  "reasoning": {
    "task_resistance": "Re-threading a broken warp end on a running loom, pulling a jam out of a knitting head without dropping stitches, and re-tying beams by hand are still hand-eye jobs a machine can't cheaply do — but pattern loading, tension setting, needle-break detection and stop-motion already run off the machine's own controller, so roughly half your shift is monitoring work the equipment does for itself, which puts it at 10 rather than 14+.",
    "embodiment": "You are on your feet on a mill floor for the whole shift among lint, 85+ dB noise and heat, reaching into moving needle beds and creels and doffing rolls that weigh more than you want to lift — 13 rather than 17 because the environment is at least a fixed indoor plant with the machines in known positions, not a construction site or a customer's roof.",
    "liability_shield": "There is no state license, no board, and no certificate anyone checks before you tend a Santoni or a Picanol; if a beam runs off-quality the mill eats the scrap and the QC department writes it up, so nothing legal has to stay in a human's name for the line to run.",
    "trust_premium": "The yardage leaves the mill with a lot number, not your name on it, and the buyer at the converter has never heard of you — the only relationship that matters is with the shift supervisor and the mechanic, which is why this is 2 and not 0.",
    "judgment_accountability": "You decide when a fabric flaw is bad enough to stop the machine, when tension drift means a mechanic rather than an adjustment, and how long to run before doffing — real calls, but they're bounded by the style sheet, the standard operating procedure and the loom's own alarm thresholds, with the designer and process engineer owning anything ambiguous, so 5 rather than 9."
  },
  "rationale": "The work — threading yarn, setting machine gauges and patterns, watching for broken ends, clearing jams, logging output — resists language AI but has been the primary target of mill automation for forty years, and employment (13,030 nationally) reflects that. Modern computerized knitting and weaving machines already load patterns digitally, detect thread breaks with optical sensors, and stop themselves; the residual human role is loading, repair, and exception handling on the floor. No license, no client relationship, and pattern-setting decisions are made by designers and engineers upstream, so the only real moat is that a robot still can't cheaply re-thread a loom.",
  "outlook": "Headcount keeps shrinking through the 2030s as remaining US mills consolidate around automated equipment; the surviving jobs are machine fixers and setup techs, not tenders.",
  "what_would_raise_it": {
    "levers": [
      {
        "dimension": "embodiment",
        "change": "Re-threading a broken warp end on a running loom, doffing, and clearing lint/jams in dusty mill air remains the specific manipulation task no commercial robot cell has priced out; if remaining US mills continue shifting toward technical textiles (aramid, carbon fiber, medical narrow fabrics) where lot sizes are small and changeovers frequent, the day becomes mostly setup and repair rather than tending, raising the physical share",
        "plausibility": "already happening",
        "would_add": 3
      },
      {
        "dimension": "task_resistance",
        "change": "Genuine two-tier split: sensor-driven stop/restart is already automated, leaving diagnosis of recurring defects (tension mapping, needle/sinker wear, yarn lot variation) as the residual. Where a mill runs 20+ machines per tender, the job converges on troubleshooting technician work — visible in job postings retitled 'knitting technician' with mechanic-level pay",
        "plausibility": "plausible",
        "would_add": 3
      },
      {
        "dimension": "liability_shield",
        "change": "Defense-related sourcing (Berry Amendment) and medical/PPE textile certification already require documented in-process inspection; if AS9100 or FDA QSR-style audit rules for implantable/surgical textiles are read to require a named human operator sign-off on each lot's loom setup record, a traceable human signature becomes mandatory. Narrow, and only for the fraction of mills in those markets",
        "plausibility": "plausible",
        "would_add": 3
      },
      {
        "dimension": "judgment_accountability",
        "change": "On short-run technical textile jobs where a single defective roll of ballistic or medical fabric is scrapped at high cost, the decision to stop the machine and scrap versus run is pushed to the operator rather than an upstream engineer; formalized as documented stop-work authority in a quality system, this is a real consequential call",
        "plausibility": "plausible",
        "would_add": 3
      }
    ],
    "ceiling_note": "No plausible route to trust premium — buyers of woven and knit goods specify fiber, construction and price, and cannot identify the operator. Employment has fallen for forty years through capital substitution, not language AI, so nothing on this list reverses the headcount trend; the levers describe a smaller, more technical residual job, not a larger one."
  },
  "adjudication": null,
  "employment_history": {
    "points": [
      {
        "y": 2017,
        "emp": 20920,
        "wage": 27960
      },
      {
        "y": 2018,
        "emp": 21190,
        "wage": 29160
      },
      {
        "y": 2019,
        "emp": 21130,
        "wage": 29980
      },
      {
        "y": 2020,
        "emp": 18830,
        "wage": 31090
      },
      {
        "y": 2021,
        "emp": 16330,
        "wage": 33990
      },
      {
        "y": 2022,
        "emp": 16900,
        "wage": 35750
      },
      {
        "y": 2023,
        "emp": 15980,
        "wage": 37130
      },
      {
        "y": 2024,
        "emp": 14530,
        "wage": 38260
      },
      {
        "y": 2025,
        "emp": 13030,
        "wage": 39530
      }
    ],
    "from": 2017,
    "to": 2025,
    "change_pct": -37.7,
    "comparable_from": 2019,
    "spans_soc_revision": true
  },
  "pivots": [
    {
      "slug": "cleaning-washing-and-metal-pickling-equipment-operators-and-tenders",
      "title": "Cleaning, Washing, and Metal Pickling Equipment Operators and Tenders",
      "verdict": "EXPOSED",
      "risk_resistance": 42,
      "median_wage": 43530,
      "overlap": 87,
      "skills_to_close": [
        "Social Perceptiveness"
      ]
    },
    {
      "slug": "refractory-materials-repairers-except-brickmasons",
      "title": "Refractory Materials Repairers, Except Brickmasons",
      "verdict": "EXPOSED",
      "risk_resistance": 65,
      "median_wage": 61290,
      "overlap": 86,
      "skills_to_close": [
        "Repairing",
        "Coordination",
        "Equipment Maintenance",
        "Social Perceptiveness"
      ]
    },
    {
      "slug": "welders-cutters-solderers-and-brazers",
      "title": "Welders, Cutters, Solderers, and Brazers",
      "verdict": "EXPOSED",
      "risk_resistance": 62,
      "median_wage": 53750,
      "overlap": 84,
      "skills_to_close": [
        "Science"
      ]
    }
  ],
  "license": "https://cookedindex.com/terms"
}