I’m not from the US, but would the initial capex and the ongoing maintenance expenses for the robots really be cheaper than a couple of minimum wage workers. IMO, a better approach would be to start with “people in trucks”, and gradually introduce robots as they scale. Something like Uber’s vision of offering cheaper rides using self driving cars (but more realistic).

Yes, though cynically, I think the idea is that venture capitalists only open their wallets if they can see your robot in action. Prudent iteration doesn't attract money.