Probably not very hard, but it will cost 10x more than regular clothes. Current process is highly efficient and produces thousands of units per day with a dozen of low skilled workers. Whatever is not automated (and it's not much) benefits from dirt cheap labour in Bangladesh or Vietnam.
Your sewing robot will make a dozen of units per day and cost hundreds of thousands. Monthly salary of just several sw engineers from SV is probably greater than total operational expenses of a medium-sized factory in India for a year.
There was a ‘startup’ that planned for the client to take a bunch of measurements and would then manufacture a garment to those. A post, which floated on HN, detailed why it turned out to not be too feasible.
Edit: it was ‘Getwear’, found the name through the design company portfolio. They specialized in jeans. Alas, since the site is dead, I can't locate the postmortem, only this article in Russian: https://vc.ru/offline/6359-getwear-close
Basically, it seems that the demand just isn't there.
Suits to measure aren't a new thing, but suits are expected to last for quite a while and thus they cost a bunch. Additionally, dunno about Indochino's web offering but they have offline shops where somebody takes measurements for you the old way, so apparently that's the experience many customers expect.
Tailored clothing can be surprisingly inexpensive, especially outside the US. A 3D scan / in-person fitting and then overseas garment manufacturing could be a solid business model.
Your sewing robot will make a dozen of units per day and cost hundreds of thousands. Monthly salary of just several sw engineers from SV is probably greater than total operational expenses of a medium-sized factory in India for a year.