The later is a much easier problem.
The later is a much easier problem.
They purchased an AutoStore, then reverse engineered it, made a few changes, and claimed it as their own invention.
A system that works well with 15 robots will often fall apart if scaled up to 150 or 1500 robots. Reliability, planning algorithm complexity, radio performance, all sorts of issues start to come up.
That’s why Hatteland patented the autostore tech in ~1995 and by the time the patents expired they only supported ~100 robots.
It’s not always easy to appreciate, because everyone publicises when they install a new automation system, but nobody publicises it if they scrap it 18 months later. Being discreet about it is better for the share price.
Of course there’s still a perfectly good market for less scalable automation; grocery just has crazy financials.