Katie Canales from Business Insider just did a nice job of summarizing some of the growing pains that automated fast food restaurants are having in the Bay Area:
In July 2019, Eatsa shuttered its San Francisco locations after the company was found to be thousands behind in unpaid rent. It’s since pivoted to restaurant tech and software and rebranded itself as Brightloom, with a new deal with Starbucks announced in 2017. Just this week, CafeX closed its San Francisco locations, though its stations at San Francisco International Airport and San Jose Airport are still open. And Zume, the Softbank-backed startup known for its pizza-making robots, shuttered its pizza business and pivoted into food-truck tech and services in November 2019. Then, in January 2020, the company laid off 400 employees to further cut costs, as reported by Business Insider’s Megan Hernbroth.
Business Insider: Some of San Francisco’s robot-run restaurants are failing. It could simply be that we still want to be served by humans, not machines.
The article includes some pointers to what it is like eating in various automated fast food restaurants in case you are curious.
The conclusion of this piece is that “we’re simply not ready to be served by robots in lieu of humans.” I agree with the not yet part of this statement, but it’s less that we’re not ready and more likely that the robots are not quite ready to serve us. Automated self-service is something that is inevitable. In-ev-i-ta-ble.
The problem with most failed instances of automation, at least from an operations perspective, is that they try to simply automate the existing manual processes. Drive-through restaurants succeeded by flipping the business processes around so that a new mode of service became possible that was centered around customers visiting the restaurant through tiny hole in the wall. Automation without revamping the flow of how things get done generally just doesn’t work well. You end up with Rosie the Robot vacuuming rather than a Roomba.
We are still in the early years of automated fast food. What most people don’t realize is that ATM technology came out in the late sixties and it took over a decade of tweaks and innovations before it reached the point where it became ubiquitous. The same will happen with fast food. The current problems are just bumps along the way and it’s no coincident that these problems are happening in the Bay Area, where a lot of the innovation in automated self-service platforms is happening.
I’m not saying that people don’t care about having relationships with people, just to be clear. But the kind of transactions we have at most fast food places can’t really be said to be relationships. Automation is going to have a much tougher go in regular restaurants. There, it will likely focus more on back-end prep, cooking, and cleaning technologies. To automate the customer-facing part of a restaurant, it’s going to require a flip in the way we think about service there.
When one watches customers at the kiosks in McDonald, trying to figure out what button to push, how to customize a burger, or how to pay, while meanwhile those of us in the line are dealing with cheery people who take our order in half the time, it does show that they have a way to go to make the kiosk faster and easier. Even the new chip card technology needs improvement, as virtually every store’s terminal operates differently from the next store. Pull out the card, no, leave it in, press the green button for amount ok, press Yes or No for amount. But, as you cite the development of the ATM machines, one day we will think it odd to see a person at the counter!
That’s absolutely true, Bill. Automation in the factory was hard to make user friendly, but there weren’t that many people working with those systems and so the answer was simply to train the workers. That works in manufacturing, but not in services. Here, the bar is much higher because the automation has to work for end users and there are simply too many of them to think about training.
You raise a good point about the lack of consistency between these user interfaces. That’s part of what an operating system does. Hmmm…
Distressing but one of the most probable things to come. Of course, our human ability to habituate to anything will eventually serve in automation’s favor.
Thanks for dropping by, Ragini. That’s an interesting point. For this stuff to work, we need to agree to it. I suppose that strong advocates of market forces would argue that if there is a demand for human touch, there will be supply. But sometimes, as with Amazon, the competitive forces make offering that touch come at a price premium.