Will Robots Take Over Our Restaurants?
Tau Ventures is an applied AI fund in Silicon Valley and food automation is one of our core focuses including an investment in Blended. Our view in Nov 2019 was that capital is plentiful, automation won’t replace the humans and that the real savings come from economies of scale. Since then the industry has been shaken in huge ways. There have been some major entrants, for instance Uber founder Travis Kalanick’s CloudKitchens became more public (supposedly raising $400M). They faced some major failures, for instance, Zuma pizza which had raised $375M from Softbank alone essentially imploded. And biggest of all, a global pandemic has ravaged the restaurant industry.
So robots will take over our foods? This blog post will bring the core benefits of these robotics.
1. Robots = Production Yes, Distribution Yes, Service No
Robots today are now performing well like humans at most of the tasks, they can chop vegetables, peel potatoes, flip burgers and terrible at operating in unconstrained environments. You could even give them huge amounts of training data to expand their functionalities, but still in a limited environment. Being able to take care of a patron fully, like a human does is still eons away. Rajat Bhageria from Chef Robotics has a provocative recent post in TechCrunch discussing in more depth the subject on narrow versus broad AI in food automation for those interested in exploring further.
At Tau we believe automation will increasingly help the production, especially the mass production of food. Also distribution, whether it’s akin to a vending machine or getting the food fully into your home through point-to-point delivery. Production and distribution together often add up to more than half of the total cost, and the pandemic has highlighted the need for traceability and supply chain visibility.
What automation can help the least is service since a large part of the appeal of a sit-down restaurant is the overall experience, including the ambiance. We are taking specifically about mid-high end restaurants, not fast casual where we see ghost kitchens with office delivery becoming more prevalent. Our view is the sit-down model is largely on pause right now will invariably make a comeback, even if automation can’t help it significantly.
2. More Startups Going Beyond Cost Reduction
Per Pitchbook, in 2019 the food tech world broadly claimed $13.5B in venture capital across 440 deals worldwide. 2020 is surely going to be the year of significant downturn, with less deals and less capital, but at Tau we believe that relative proportion of food automation will increase and some of that will sustain much forward. In general we don’t see the pandemic as having overwhelming created new circumstances but more so accelerated secular trends. Reducing costs is always a driver behind automation and remains as such but add to it a desire for more
· Safety: Contactless preparation reduces contamination risk, there is also the potential to improve safety in supply chains.
· Convenience: Ghost kitchen facilitates delivery.
· Innovation: For instance Minerva uses different raw materials and novel processes for making tortillas.
3. Consumer Automation Is Very Far
Our view is that food automation is still years away from truly permeating our homes. In computing we had mainframes and minicomputers for decades before personal computers became significant enough. What partly drove that adoption was that PC’s fundamentally made people far more productive, in fact allowed us to do tasks that weren’t even possible before. Not the same case with food automation, indeed consider the deep traditions embedded into cooking and how it relates individuals, families and communities. Yes we have some automation, from fridges to ovens and microwaves to mixer, but a far cry from the level of automation in enterprise. In the home kitchen, its’ indeed not all about costs and chores.
Previously industrial robotics would reach out to factories, now they are starting to reach out to restaurants providing mostly white-labeled and cobranding models. A handful may succeed in creating their own brands but those will be very capital intensive. So when we look for investments in food automation we are asking immediate questions around CAC and LTV but also around how fundamentally the business scales, especially around unit economics.
