Does the hanger a sim loop designs change for stretchy, heavy knits, and what does a tee-designed hanger cost them?

What it found

Runs hanger-fit's CMA-ES and random-search design loop (150 scores, 3 seeds) on five size M cases taken from hanger-sweep's corners: base tee, stiff (EA 800), soft (EA 40), soft and double weight, and soft, double weight on a 10 cm strip. The best arm half-width is 0.9 cm past the half shoulder for the tee and 1.5–1.8 cm for the stretchy cases, but slope, bow and end radius scatter between seeds at similar error, so only the width is identified. A tee-designed hanger scores 0.22 cm RMS on the tee and 1.06 cm on the soft heavy knit; one designed for the knit scores 0.52 and 0.76 cm. CMA-ES beats random search in all five cases. The narrow-strip case is poor for every hanger (best 0.99 cm). Fabric constants are swept corners, not measurements; 2D model, one size, shape error only.

A figure from its own output

Paper, code and results

• README and method
• Experiments and results
• Code
• Tests

More in Laundry-hanging robots

• Should a laundry-hanging arm also steer clear of the other arm's next pick? A paired test on 1000 loadsShould a laundry-hanging arm also steer clear of the other arm's next pick? No: paired on 1000 loads, lookahead loses to avoiding the current pick in 14 of 18 settings and ties in 4, by up to 11 an hour. Stylised: inputs are placeholders.
• What does it cost two laundry-hanging arms to empty the same dryer pile?What does it cost two laundry-hanging arms to empty one dryer pile and wait for each other? Blocking at 0.30 m clearance takes aiming from 201 to 191 garments an hour and blind pinching from 179 to 151; a third arm adds only 3 to 5. Stylised: clearance and spacing assumed.
• Can a laundry hanger be printed in two halves on a 220 mm bed, and is the bolted joint strong enough?Can a laundry hanger be printed in two halves on a 220 mm bed, and is the bolted joint strong enough? A half-lap hub joint with two M4 bolts lets every width up to XXL print on a 220 mm bed; the joint (safety factor 3.7–13) is then weaker than the arm (40–70), and printed pegs fail. Stylised, assumed material constants; nothing printed.
• Does the hanger-fit width result survive when its assumed fabric constants are swept two orders of magnitude?Does hanger-fit's "arm 1 cm wider than the half shoulder" survive 162 swept fabrics (bending rigidity 0.003–1, weight ×0.5–2, stretch 40–800 N/m, strip depth 10–30 cm)? Best offset is +1 cm in 151 and +2 cm in 11, all of them stretch 40 N/m; bending rigidity alone moves the best error under 10%. Constants swept, not sourced.
• Does a second 3D cloth model, wrapped over the hanger arm, agree with the 2D hanger drape?Does a second 3D cloth model, wrapped over the arm, agree with the 2D hanger drape? On the cases both run, it picks the same best arm width (22.8 cm for a size M tee; error correlation 0.994) and adds the bend across the arm (119 deg/cm over a 0.5 cm arm, 34 over a 2 cm one), but disagrees for a heavy knit on a thin arm and is not grid-converged. Stylised.
• Should a laundry-hanging arm yield to the other arm's next pick only when the clash would be long? Gating lookahead by predicted overlap, paired on 1000 loadsShould an arm yield to the other arm's next pick only when the clash would be long? Gating by predicted overlap removes most of lookahead's loss and gains +0.7 to +1.35 garments an hour for pinching at 0.30 m (paired, 1000 loads). Stylised: inputs are placeholders.

All 25 Laundry-hanging robots projects →

Other research topics

Proper scoring & elicitation28Effort & peer prediction9Verification games & slashing19Detection & tolerances14Markets as routers13Prediction-market design9Paying contributors8Decentralised training systems23Digital twins & sim-to-real164LLM chip design loops40Robot manufacturing loops53Ambient AI assistant46Self-replicating manufacturing34Simulation-first design64Engineered clothing45Task-specific AI accelerators40Algorithmic chip-design loops42Open-source purpose-built robots44The software-engineering loop43Adversarial human imitation41Open-source residential construction41Autonomous multi-angle capture34Sensing, imaging and 3D31Energy supply and dense energy storage39Robot societies among people31Agentic co-design of low-cost robots27Laundry-hanging robots25Machine learning on qualitative data23Robots on the building site27Civilizations & group psychology1History & human sciences1

Produced with AI assistance (Claude). Results are from the code in this repository and are simulated or stylised unless the abstract says otherwise; nothing here implies affiliation with or endorsement by any lab, researcher or company cited. MIT licensed.