RaaS, Restaurants and Cooking the Perfect French Fry

It was only a matter of time before the robot-as-a-service model made automation viable for sectors other than manufacturing.

Flippy 2, engineered by Miso Robotics, prepares fried foods like French fries. Image: Miso Robotics

Flippy 2, engineered by Miso Robotics, prepares fried foods like French fries. Image: Miso Robotics

Cobots and robots may soon start doing many of the essential duties of fast food and other restaurant workers. Their contribution will change the nature of the 500,000 jobs the foodservice industry workforce expects to add this year and the 15.5 million jobs in the industry overall. The fast food sector offers many opportunities to leverage machine labor since common tasks such as frying are tedious and dangerous.

Fast food is the frontier for the integration of robotics into foodservice because it has the lowest prices for dining establishments. The rise in costs for food, gas and labor constantly motivate owners to find new ways to reduce expenses.

One robotics engineering companies designing solutions that meet the needs of fast food restaurants is Miso Robotics, headquartered in Pasadena, Calif. Miso currently produces four robotic solutions: Flippy, a modular frying robot, Chippy, a frying and finishing robot, CookRight, AI-powered sensor systems, and Sippy, an automatic beverage dispenser. Here the term “finishing” refers to putting the finishing touches on a dish, like heating a grilled cheese sandwich to ensure the cheese is melted evenly.

Buck Jordan, Founder, President, and Chairman of Miso Robotics, makes machines that are standard, automated and user-friendly. He strives for robust designs that require minimal interaction with kitchen staff.

“The less kitchen staff need to interact with robots, the more they appreciate the technology. Ultimately this is what automation in food service must do: make staff’s jobs safer and easier—and if done exceptionally well—friendlier,” says Jordan.

The basics of designing a foodservice robot

Fast food restaurants tend to use robots rather than cobots because the machines do not need to be located near or interact with people in a shared space. As an example, Flippy sits in a dedicated fryer area handling between two to five fryers. There is a safety barrier with an automatic emergency stop between Flippy’s robotic arm and human workers.

Flippy is not considered a cobot, although Miso complies with cobot standards. Flippy moves quickly, up to 1.5 meters per second and can handle payloads up to eight kilograms. Workers all over the world know the risks of working too close to a deep fryer:heat, slipperiness, messiness, the risk of burns, and smelling like oil. Having a robot handle the frying not only eliminates these risks but also reduces overcooking, waste and cross-contamination.  

Miso’s machines take up a minimal amount of space. Currently, Flippy requires seven inches of aisle space. Miso’s engineers are working to reduce that further, keeping in mind safety and Americans with Disabilities Act (ADA) requirements.

Miso’s engineers put a great deal of focus on its articulating arm which contains a series of modular motors that could be susceptible to clogging, damage and debris generated from the cooking process. Miso engineers have redesigned this component numerous times to eliminate exposed wiring and joints. During cooking, the arm is covered with a special sleeve.

The robotic sits on a rail which moves left-to-right or right-to-left, depending on the workflow of the restaurant. Flippy’s other components include a universal frame, which fits over the restaurant’s existing fryers; a safety barrier; and an AutoBin and specialty basket assembly, which can be located on either side of the fryers.

Another task for engineers is to simplify the process to clean robots.. Flippy has an on-board touchscreen that guides the staff through an easy daily cleaning process. Staff can clean Flippy’s less sensitive surfaces in the same way that they would wipe down any other National Sanitation Foundation Institute-certified (NSF) surfaces in the kitchen.

Making these types of robots easy to understand and repair is critical, especially considering that workers in the service industry aren’t often exposed to this type of technology. The employee training time for Flippy is typically less than one shift. Flippy’s arm is designed so a technician can access and replace the modular motors without specialized tools.

Other goals include ensuring the machines can withstand harsh conditions, like high heat and humidity. And they need to run continuously and remain reliable. Flippy is able to operate for hours through lunch and dinner shifts. Miso uses Robot Operating System (ROS) on Linux to help build the robotic controls.

The testing and validation process

Testing a foodservice robot involves virtual simulations. Miso gathers information about customers’ fryer dimensions. It leverages Amazon Web Services (AWS) RoboMaker to run simulations prior to installation.

“Miso tries to be fryer brand agnostic. (This makes) it easier for restaurants to incorporate automation onto their existing equipment,” says Jordan.

Further, the testing process requires making large batches of food that the robot will regularly prepare. This prep work occurs in Miso’s lab. For Flippy, the robot’s AutoBin AI cameras identify foods placed in the AutoBins. Flippy then uses Machine Learning to recalculate the most efficient order of operations every few milliseconds. The combination of technologies ensures foods are always cooked to specification.

The data that Miso collects are helpful for business intelligence—digital transformation is not just reserved for manufacturers. With regular data collection, a fast-food company can analyze information regarding food quality, supply. inventory and speed of service. 

For example, certain items, like French fries, are easier to cook than others.

“Fries, especially for larger brands, tend to have great consistency. Once Flippy has learned how to cook them, it gets (them) right more often than human staff. Regarding oil, Flippy provides precise data back to the operation about what (items) and how much (of them have) been fried in every slot,” says Jordan.

Over time, Flippy can make better decisions about how much oil to use and when to discard oil.

Miso currently offers machines like Flippy as Robotics as a Service (RaaS). This is one of the first instances of the phenomenon outside of manufacturing. Restaurant owners incur a lower initial capital expense than they would if they purchased the robots.

Jordan says his tip for engineers designing robots for fast food restaurants is to listen to what the customer needs a robot to do. The engineer is supposed to design a machine that will make a restaurant team member’s job easier.

“Typically, the minimum staffing for a fast-food restaurant is three people. No amount of automation will get rid of those three people. They need to be there to manage customers and handle unpredictable events,” says Jordan.

With the right robot, a restaurant can improve the food’s quality, consistency, and the speed of service. Reducing customer wait time is important because it keeps customers happy and allows more food to be sold. Jordan says the “secret sauce” for a fast-food restaurant is to serve more products and earn more revenue.

Miso Robotics is now trying to determine whether fast food restaurants should fully automate food and drink preparation.

““For example, should we take out the fryer and the drink station? Should we ask our restaurants for input to build a series of self-contained robot kiosks? If we did that, how would employees’ jobs evolve?” says Jordan.