HAX’s VC-Founder Industry Hard Tech Matchup is now accepting applications. Register today!
Back to news articles
Lessons on automation from our investment in xRobotics, and notes on why Zume failed
HAX Team
xPizza One is an automated pizza making machine
xPizza One is an automated pizza making machine. Source: xRobotics

xRobotics (HAX 2019) just announced that it’s launching an automated pizza robot next week, and we have a few insights from our hands-on collaboration with them on why robotics and automation can be really challenging to get right, especially for early stage startups.

When Zume killed its pizza automation business at the beginning of 2020, many thought Pizza robotics, and food automation in general, were dead. We had just invested in xRobotics a few months prior. At the time they were a small team based in SF/Moscow that was working on the same problem, but with a solution streamlined to better address the challenge, that had the power to revolutionize almost any pizza parlor in the world from day one.

Their first product, the xPizza One, is a small form factor pizza automation station.

At every turn, the design leverages the strengths of low-cost robotics-completing repetitive, non-predictive tasks, in fixed environments-and the constraints of restaurant owners.

A restaurant can’t run the risk of down-time in the kitchen so the team designed the xPizza One to be cleaned in 15 minutes and modular at every level. If any individual part breaks, the entire system can keep running at full speed. It’s also the same size of the ingredient storage fridge that pizza parlors everywhere use, meaning any pizza parlor in the world could be using this system tomorrow in their existing kitchens.

They’re handling 23+ toppings and 5+ sauces at launch with a max capacity of 150 pies an hour (though standard pizza throughput is around 60). They’re also leaving the tasks that humans are good at, to humans. Humans still roll out the dough, interface with customers, chop ingredients, and keeping the business organized and operating.

So why does HAX think xRobotics will succeed where Zume failed?

If you take a look at Zume’s core processes, they tried to mimic how a human would make a pizza, rather than optimizing for what robots are good at. Spoons and paddles were attached to Robotic arms that were designed to spread sauce onto the dough and lift the pizzas in and out of the oven.

Another key challenge was that they were using equipment designed for incredibly precise manufacturing processes to do it. These systems are not only expensive (the 6 Degree-of-Freedom robot and control systems likely cost six figures) but they take up a large amount of space, something pizza parlors don’t normally have the luxury of doing, as a majority are small real estate operations with very limited in store dining.

Zume never publicly announced a foray into topping automation – the step which takes up 95% of employee time. I am curious to know how they would’ve tackled it, but the steps they did automate (applying sauce to the pie, and placing a pizza in the oven) just aren’t time consuming enough.

Most parlors are also heavily skewed towards delivery and takeout, so a flashy robot making your pizza isn’t really a selling point. Dodo Pizza, one of xRobotics’ first partner organizations, did >70% takeout/delivery last week. Their focus is on getting quality pizzas to their customers in a fast and consistent manner, and every design and product choice by the xRobotics team helps them achieve this.

xRobotics is employing a HaaS model – pizza chains pay a down payment for the machine and a recurring fee to use the system, which includes 24/7 technical support worldwide.

Their first robot will be installed at the Dodo Pizza in Oxford, Mississippi (right next to the Ole Miss campus). They’ll be in 50+ locations across the USA in early 2021, and have more than 400 locations in the pipeline for EOY 2021.

See Forbes’ coverage of the xRobotics launch here.