The robots are among us. They have taken over our institutions, our cities, and our homes. Sounds a little like the plot to a Terminator movie, doesn’t it?
But in fact, robots are among us, and lately have been employed by quick-service franchises across the country. For example, it's not just Johnny who will be cooking your burger but a robot named Flippy, who was created by Miso Robotics and is in the employ of one unit of the CaliBurger chain, with plans to be in more by 2018.
Miso Robotics is a startup that came to life last year when CEO and cofounder David Zito met with John Miller, chairman of CaliGroup (which owns CaliBurger). “We came up with the idea together to leverage [artificial intelligence] to bring industrial automation tools into a commercialized setting,” Zito says. “In this case, the back of house of restaurants. The idea here is to bring in a kitchen assistant to aid the overworked kitchen manager, chef, or back-of-house employee so that they can do the things that humans are really better at like hospitality, menu planning, and cleanliness.”
Flippy's job is to make burgers. It is a robot arm connected to cameras and sensors that can sense the volume of food on the grill, the temperature of the grill, and the temperature of the food it is cooking (burgers for now, chicken in the future). The concept is to help increase kitchen productivity notably during peak times. “We're targeting a 50 percent increase over what John is currently experiencing,” Zito says. “An average grill cook can do about 200 [burgers] an hour. We're targeting 300 an hour.”
Miller says his company has considered various kitchen challenges, such as the tedium, heat levels, and potential burns for grill chefs, and how those could be aided by technology. At the front of house, CaliBurger invested in giant FunWall video displays for their restaurants where customers use their mobile devices to interact and play video games. How do these two aspects connect? The goal is to free up human staff to engage more with customers, and an optimal way to do that is via increased kitchen automation.
The initial Flippy tests were done by placing a lab next to its Pasadena restaurant, allowing for “this nice, cooperative development and testing cycle,” Miller says. “We can develop things in the lab and quickly test them in the restaurant. We have cameras in the restaurant that are used in the lab, and everybody goes back and forth. I think that's really the only to effectively test something like this.”
Miso Robotics plans to ship a pilot unit early next year, and once it has been well tested they plan to bring them into all of the CaliBurger locations, which should number 50 by the end of 2017. CaliBurger is thrilled about the prospect, Miller says. “We're setting up our Pasadena store so we can deploy Flippy this summer to make burgers for the public. Our goal is as quickly as possible to serve people their first burgers made by a robot.”
The real vision behind Miso Robotics is to go beyond Flippy. “We are leveraging [artificial intelligence] to build a cooking platform for the future,” Zito says. “We can control industrial arms in a commercial setting at price points that can tackle lots of high-value tasks in the kitchen—prep, frying, stirring, boiling. The imagination's the limit. We hope eventually to be able to train our AI to help step in and handle tasks when a kitchen assistant is needed.”
Zito says hat Flippy fits into a standard kitchen galley and adapts to standard kitchen equipment. He says their robot “does not require a kitchen owner to outlay a big model and [they can] take advantage of our platform and bring it to their bottom line.”
Switching focus from food creation to delivery, San Francisco-based tech firm Marble has been testing out food delivery robots in the Mission and Potrero Hill neighborhoods since the beginning of the year, and the company's first partner is Yelp Eat24. Harrison Shih, founding head of product and operations for Marble, says they have gotten mixed reactions from people but a vast majority have been positive, especially from children. “We have chaperones with the robots who get stopped and asked a lot of questions about what it's doing and how the technology works,” Shih says.
The robots, which could also be used to deliver groceries and pharmaceuticals, are being built with a technology that they call “hybrid autonomy,” The machines will mostly be autonomous but will always have a human in the loop.
“As it's driving down the sidewalk and hits an obstacle that it doesn't know how to get around, it will phone home and ask for help,” Shih says. “Crosswalks can be tricky sometimes. For right now and for the foreseeable future, we do have a person walking alongside of it, and this person is there mostly to collect data [to see] how it's interacting with the world. It's a chaperone for the robot as we're still testing out there.”
The tests are being done at walking speeds and at short enough distances for customers to be able to get their food within 30—35 minutes. As Shih notes, people like their food to be delivered in an efficient and cost-effective manner. While some customers have issues with coming down to the curb to get their food (they enter a 4-digit PIN to open a panel and access their delivery), with “more than 90 percent of our orders people have articulated that they would be happy to come down and meet a robot and would definitely do a robot delivery again,” Shih says. “We take that as a very positive sign.”
Observing an increased trend with consumers for home food delivery, Shih believes Marble's technology “will bolster the restaurant and food industry. Some people still want to go dine in, but now more and more people want the option of getting prepared meals delivered straight to them quickly and efficiently.” Such efficiency includes sending confirmation texts with ETA info and the necessary PIN number to customers. There are other possible interactions depending upon which establishment one orders from through Yelp Eat24.
Shih says these rolling rectangular robots can be complementary to human delivery. Shorter runs are easier to assign to robots, and tips for human employees might be smaller compared with farther runs, not to mention easing the issues of double parking and traffic congestion. For customers who might pick up for shorter deliveries they might accept it from a robot. With more delivery options possible, businesses could increase sales more cost effectively and efficiently.
“They will be able to sell more product, help the neighborhood, and people will be able to experience and enjoy their neighborhood favorites with less challenge,” Shih says. “One of our key focuses right now is to bolster the neighborhoods.”
An interesting invention that is being used in bars and full-serve establishments, which will apply more to quick serve in the future, is the Nectar Labs invention called the Connected Pourer and Stopper. Aayush Phumbhra, co-founder and CEO of Nectar, says this connected device was originally created for milk jug monitoring (how many times have parents run out of milk for their baby?), but when he and his co-founder (COO Prabhanjan “PJ” Gurumohan) looked into the alcohol market and did their research, they saw the potential there.
Having spent time in restaurants and at bars, Phumbhra knew “there was a huge problem in the alcohol market with inventory management and shrinkage,” he says, adding that he believes there is no other technology that exists to help bartenders keep track of inventory and know when to re-order. As he points out, alcohol is money in a bottle. For restaurants and obviously bars, “liquor is the highest cost and highest profit product for them.” A recent Nectar press release claims that heavy pours and outdated inventory methods are costing the global liquor industry more than $50 billion per year, and they know how to stem that tide.
The Nectar Connected Pourer and Stopper, which the company says emerged from 50,000 hours of R&D, starts at $299 per month. It allows bartenders and restaurant managers to keep an inventory 24/7 in real-time with each pour, with data sent to an app that is used for tracking and collecting data. “We keep track of bottle level in real-time which means in effect we also know how much was poured out as we knew the level prior and the level after the pour,” Phumbhra says. The connected device also keeps track of when bottles fall below an ounce and allows for timely one touch re-orders from distributors. Automating these manual processes cuts down on wasted man-hours after closing and on holidays, and they are understandably touted as being more accurate than pen-and-paper tabulations.
How the Connected Pourer and Stopper collects data is simple: A Bluetooth cap is placed on each bottle; the basic idea is to keep a cap paired with a specific brand and size of bottle. One can re-pair a cap with a different brand, but so far Phumbhra has not heard of any clients doing that. The cap setup is easy and takes 10 seconds. The UPC code is scanned once, and then the bottle is ready to be monitored. “We made the process super simple,” Phumbhra says. “[People] are able to set it up very, very quickly and start using it right away.”
Nectar has been doing private pilots and rollouts of the Connected Pourer and Stopper for the last two years. They began with a 3D printing unit for testing, and for the past nine months their final assembled units have been rolling out in San Francisco and Miami. Their plan is to make it available to all U.S. cities by the end of the year and in other countries by 2018. They feel the problem of liquor shrinkage is not just national but global, giving them huge market potential.
“One thing we hear from bar owners today is that it looks so simple that they can't even tell the difference,” Phumbhra says. “That was our main goal. We didn't want drink technology at the bar that looks like technology. We wanted it to seamlessly integrate into the bar environment so no one really thinks about using technology. I think that's the main problem where people try to put in too much technology and change behavior—it never gets utilized.”
Flipping burgers, delivering food, maintaining bottle levels—how far will food and restaurant robots take us? We will find out more in the near future. As Zito jokes: “Everyone's dream is to just say, 'Hey Siri, hey Alexa, I want a burrito.' And it just drops into their hand wherever they are.” As long as that does not turn into a sinister Skynet scenario, things should work out just fine.