[leoc]

> While tailors have figured out a formula for men’s suits, bra tailoring is a younger technology with a smaller market and far fewer competitors. [...] But bras, coming after the Industrial Revolution, had no such history of custom tailoring.

Even companies trying to custom-make suits without multiple individual fittings are apparently still very much wandering in the wilderness, so it's not surprising (though of course disappointing) that an effort which started futher behind didn't succeed. Maybe this also relates to Boeing and SpaceX's parachute challenges https://news.ycombinator.com/item?id=21850831 : it seems fabric remains hard.

[Robotbeat]

Fabric is a very hard problem because it has an extremely high number of degrees of freedom. Flexible materials are like this consistently. If you want automation, with few exceptions, you want stiffness.

Tesla has been unable to automate wire harness installation because it's super flexible with many effective degrees of freedom (they hope to solve that using a stiffer flex cable that consolidates the wiring). Even moving some fiberglass fluff with a robot was an unreliable bottleneck they eventually removed.

I even saw a project once that attempted to automate cloth handling by first stiffening the cloth with starch so it could be more predictably moved from place to place.

Cloth is fundamentally hard. It's not impossible to solve these problems, but it's not at all trivial.

[jiggawatts]

> first stiffening the cloth with starch so it could be more predictably moved from place to place.

Hah! I thought I was the only one to have this notion...

I was thinking of impregnating cloth with a hard wax similar to candle wax and then melting it out after automated sewing.

Starch makes more sense from a cleanup perspective, but it may not have the required control over stiffness or other useful properties.

For example, wax would allow temporary joins using pressing and heating, which would then hold the material in place before the permanent sewing.

[remarkEon]

Interesting observations. For some reason I’m reminded of this (admittedly pretty bizarre) concept car from about a decade ago by BMW.

https://youtu.be/0pwabDeqVi8

[floatingatoll]

The BMW Gina was a direct-to-museum car that used 'polyurethane-coated Lycra' wrapped around a hydraulic (modifiable) frame.

Previously on HN in 2008, and a bonus Wired link:

https://news.ycombinator.com/item?id=214314

https://www.wired.com/2008/06/bmw-builds-a-ca/

[remarkEon]

So cool. Now I want to see one of these in real life. Thanks!

[maire]

It is more than just the age of tailoring vs undergarment making that is the problem. If tailoring was the only issue, we would just tailor women's undergarments and declare success. The real issue is that women's bodies are far more variable than men's bodies. Most garment designing software gets around this by fitting exactly the hardest part of the body and then scaling everything else. Bras are one of the more difficult parts of the body to fit although the hip area (pants) are also difficult to fit for the female form.

Pants actually illustrates the problem in a way that everyone (men as well as women) would understand. You can't just take a few measurements and expect success. You have to know how the measurement is distributed. Is the added inches on the stomach? Is it on the hips? Is it on the posterior? I won't go into the many crotch measurements - but they are also important.

[floatingatoll]

Elsewhere on HN today:

https://news.ycombinator.com/item?id=21876063

> Uniqlo heads towards full warehouse automation with groundbreaking robot that can fold and box clothes

[mynegation]

I buy my dress shirts at Indochino. They do multiple first measurements on you the first time, then they make a first iteration shirt. You come for a second fitting with that shirt when they look how it sits and re-adjust the measurements which are stored in your profile and all subsequent shirts are made to them. The result is pretty good. Never been to a custom tailor but I guess this is a good industrialized approximation of what they do. I am wondering if something like this could have worked.

[ahaferburg]

Ever since I came across this video https://youtu.be/gy5g33S0Gzo?t=4 of a robot taking 1,5 hours to fold five towels, with mixed results, I think about how much of our brain might be devoted to handling fabric every time I have a towel in my hands. Probably the same with handling fluids, like predicting how milk is going to behave in an opaque carton.