IOT in Action at Chick-fil-A
The Internet of Things or IOT, has a big problem: it’s pretty abstract for a lot of industries. In the restaurant industry especially, there are a lot of "what ifs." But more companies are sharing their tricks or strategies to tap into the sensors, data and clouds that make up the burgeoning technology.
During the 2019 IOT Fuse Conference dedicated to exploring the Internet of Things, Chick-fil-A shared some of its tricks (and a few what ifs) as it aggressively pushes toward a truly connected restaurant.
Brian Chambers, the director of enterprise architecture and chief technologist at Chick-fil-A, laid out his IOT thought process and one of projects he oversees inside the restaurants. He said it all starts at version “.1,” when companies just need to do something. To get a good start, some kind of data has to be flowing so the use case can be identified.
“Just do something. Do something that’s tied to a business need,” said Chambers. “We did some of this, we did some of this early by connecting some of our equipment to cloud providers. My only issue is we didn’t do more of this. There is business value in doing things quickly. Then, you can start building out things you need.”
Version 1.0 is collecting and analyzing data. He used the example of monitoring a fryer’s cleaning and cooking cycles. By monitoring that, the company is able to make informed decisions about how to optimize the restaurant.
“Maybe there needs to be another fryer or maybe they cook in batches different to optimize the usage,” said Chambers, saying version 1.0 is all about “gathering that data and doing it in a coherent manner so you can make that decision.”
Version 2.0, where Chambers sees the company now, is measuring, sharing and alerting. He gave the example of hot holding chicken in the back of house.
“There is laser etching around the pan, it’s a barcode. When the chicken is finished it goes in that pan and it’ll be put on the bracket," said Chambers, describing the scanning apparatus. “And this 3-D camera is going to read the edges of that pan and see that it’s spicy pan No. 3, and it will immediately start at timer so we can see that the time is running down on this pan.”
That timer is recorded, sent to a screen to alert local staff but also to the restaurant’s on-site computer and gateway to the cloud. That local computer also does “edge computing” or computing out in the field instead of on the cloud—important for aggregating the data and as a failsafe if internet connectivity goes down. From there, the data is sent to company headquarters, where further analysis could monitor waste, throughput or opportunities for optimization.
Version 2.0 requires a few things, but they all have to work in tandem from the pan to the scanner and through the edge computer (a cheap commodity device) and the cloud connection.
“It’s a very simple solution but it requires various solutions,” said Chambers.
And then there’s version 3.0, autonomous decision making. And be warned, this is where the “what ifs” come in. Chambers said they’re not there yet, but this is where the industry is heading. He looked to the waffle fry station for a potential use case. It’s a rough job; he said team members ranked it No. 3 in terms of difficulty within the restaurant (he didn’t mention No. 1 or 2).
“If you are the team member working the waffle fry station, you’re going to make something like 250 decisions during your shift,” said Chambers. “We want to cook just about the right amount, we don’t want to go over and let it sit—it’s a very small hold time of about five minutes. But you also don’t want to be the person who undercooked the fries.”
It also requires handling some 4,000 pounds of potatoes per shift. It’s brutal, but with an order of waffle fries sold every 15 seconds across the 2,400 locations, it’s both necessary and a chief target for optimization.
“This is how many fries we're going to sell per minute, that’s a good start, but what if we take more inputs, like how many fries are being cooked, how many people at the point of sale and how many people are coming into the restaurant or drive thru. With all that we can paint a better picture of how many fries we need to cook,” said Chambers. “Then we have a brain application to take all that information and create a forecast, so that’s cool. Now that it knows the answer to that and potentially automate cooking that product.”
But the robots aren’t here quite yet.
“Step one, let’s just show the team member what they should do instead of looking at all those aspects, what if we just give them a screen that takes all those inputs and mashes them all together and says, ‘You need to cook 2 1/2 pounds of fries and you need to box two large and two medium orders,’” said Chambers.
And that’s the real potential of IOT, behavioral change that enhances the business operations. There’s an obvious labor saving case there, a manger that has worked the fry station for five years may get it pretty close to ideal with the least possible waste, the perfect crisp and the perfect timing. But with a truly connected restaurant, anyone could have the execution prowess of a seasoned fry guy or gal.