Ideas

Does a restaurant want a prep robot?

A trade study of eight ways to get a restaurant's prep done, from buying it pre-cut to a humanoid. Staff with good machines win; the best robot is a fixed cell that would have to cost about $80k.

DoneStarted

Building the NERO robot raised the obvious question: what would a robot like this actually be for? I picked one concrete job, prep in a single restaurant, and worked through it the way you’d scope a real product. First came a concept of operations: who does the work, when, and what any solution is judged against. Then a trade study of eight candidates, run 20,000 times with every input and weight varied.

The answer

  • Staff with modern prep machines win. They come out on top together with buying the cuts that keep pre-cut, in 96% of the runs. The machines cost about $16k and pay back in about 8 months.
  • No robot beats that on money in one restaurant. Machines already do the cutting cheaply. What’s left is trimming, loading, fetching, portioning and cleaning, spread across the day, and a robot can only absorb so many hours of it.
  • The best robot is a fixed cell at the prep station. It passes every gate and saves about 6.4 staff-hours a day, twice what the machines save. At $130k it pays back in about three years. To pay back in two, it would have to cost about $80k, or about $50k if the restaurant would otherwise buy machines.
  • The mobile robot ranks seventh.

I followed up with a detailed design of the fixed cell, interfaces first. The arm sits beside a vegetable prep machine in a well, with a chute into a pan on a scale, and pass-through drawers that double as guards.