Inside the Kitchen That Listens to Food and Cooks by Itself
I have no idea how to cook a chicken. Sure, I can defrost a chicken breast, lay it on a pan, and bake it until it won’t kill me, but that’s not really cooking a chicken. I’m talking about taking the entire body of a bird, seasoning it perfectly, trussing it, then putting it in the proper pan and placing it inside the oven for the correct amount of time, so that it tastes really, really good.
The chicken I ate at the Innit kitchen headquarters in Redwood City tasted like it was cooked by a pro. It was amazing—cooked for exactly the right amount of time and at varying temperatures so it retained flavor and juiciness without any worrying pink. But the pro that delivered this perfect meal to my plate wasn’t a master chef, but a robot.
OK, yes: There were two chefs who prepared the food—slicing, dicing, trussing—but technology did the heavy lifting, determining how hot to cook something at, when to change temperature, how long to cook these for, and how to make sure that everything was ready at the same time. Innit has created a cooking system, built on machine learning and high-tech sensors, that can detect whatever it is you’re making for dinner and produce perfect results every time. And it doesn’t just cook chicken. We also had potatoes, Gazpacho, and an apple tart, all equally delicious. For each food, it designs a cooking process optimized for your exact menu item. All you do is press play. Technically, I “cooked” our lunch: After the raw chicken went into the oven, I hit the play button, and voila.
I Smell Home Cookin’
Thus far, Innit has only created a platform that connects sensors to its software. You can’t go out and buy an oven that will sensor-bake the perfect bird for you. That is the end goal, however, and Innit is currently talking to potential partners in the kitchen appliance and food industries. They wouldn’t name any partners on the record, but the company hopes the first wave of kitchen products featuring its sensors will arrive this summer. These won’t be high-priced, rich-people ovens, either—Innit is aiming its tech squarely at the consumer market.
There are several technologies that go to work when you want to make something tasty. Innit’s software handles inventory management in your kitchen, using image recognition technology to scan the photos you take of your fridge or pantry and build an inventory list. It can then take that inventory and come up with recipes sourced from different partners to give you possible meal combos. The sensors scattered throughout your kitchen appliances can all detect different things, like when food is going bad in your fridge, or in the case of my delicious lunch, at which point to adjust the heat in the oven while baking that chicken.
Innit co-founder Eugenio Minvielle previously worked as CEO for Nestlé in Venezuela, and later Nestlé Mexico and France, eventually moving to Unilever North America. He’s spent his career managing food factories and laboratories, where great big machines, highly sophisticated software, and armies of employees were analyzing food to determine things like expiration dates, nutrition values, and organic status.
“You can get extraordinary information from food, and you can know tremendous amounts of things,” he says, “Up until now, that’s been limited to the factories and laboratories. Now we have the opportunity to bring this information to the consumer and help them answer questions.”
One question that consumers often can’t answer no matter how much data they have: “What’s for dinner?” Innit’s system recommends a wide variety of possible meals. But unlike smartphone apps that can tell you what to cook based on what’s in the fridge, Innit’s software isn’t just working from a list. Because there are sensors in your fridge that can detect gasses and odors, the system also knows what state your food’s in.
“If a tomato has been sitting there for five days, maybe it tells you. ‘Make me into a soup,’” Minvielle says. “But if it’s there for two days, ‘Make me into a salad.’”
One enabling factor that’s helping Innit move quickly, says co-founder Kevin Brown, is the ever-decreasing cost of sophisticated sensors and other hardware components. Brown is eager to place those sensors in consumer kitchens, which he feels haven’t advanced enough. “It’s a 21 trillion dollar industry, and it’s been stuck in the 1970s,” he says. “The last big innovation was the microwave, or maybe your Nespresso machine.”
Even though no consumer Innit-loaded appliances actually exist, that hasn’t stopped the team from building a test kitchen to put its platform through the ringer. The co-founders took me through the whole process. First, we threw a few veggies on a camera-rigged cutting board, which recognized them and suggested we make gazpacho. And you already know about the chicken—the perfect, super-juicy chicken. The bird was already trussed in our demo, but Innit says it will be able to serve how-to videos for whatever you’re preparing. You’ll be able to view them on a smartphone, a tablet, or a screen built directly in to a fridge or oven if that’s what Innit’s potential hardware partners would like to build.
The chicken was placed into a prototype oven (which the team described as “the self-driving car of ovens”) that uses four different kinds of heating technologies: a broiler, baking heat, convection, and a microwave—but it’s programmed to use as little of that last one as possible. The in-oven sensors detected the weight of the chicken, its current temperature, and how old it was, then processed that information to cook it precisely. It did things like lower the temperature for a while, then crank it way, way up for a few seconds to crisp the skin—a technique common in professional kitchens, but something I wouldn’t know how to do at home. I barely have the patience to cook a potato in the oven, and I usually just go for the microwave.
None of this is to say someone couldn’t wildly screw up a meal made in it Innit-enabled kitchen. For dessert, we had tarte Tatin, a beautiful, upside-down apple dessert that laid on a bed of crust. And I thought, sure, an alert could have told me those apples were prime for tarte Tatin-makin’, and an app could have walked me through the process, and an oven could have perfectly baked it, but I could have sliced the apples too think or thin, or thought, “Hell, I’ll just use this pan, even though I’m supposed to have one that’s a little bigger.” Those sort of errors can’t be entirely wiped away. But thankfully, Innit’s founders say they aren’t all too concerned with taking everything human about cooking out of the kitchen.
The team also hopes to curb food waste in homes. “The average family wastes $1,500 on expired food a year,” Minvielle says. “30 percent of food gets wasted after it goes into your kitchen. So we can help with that.”
But the company’s core goal remains: to build appliances that help us eat better, healthier, and to cook more efficiently. The smart platform isn’t designed to do everything for you, but it will help you cook food more exactly. Because, as Minvielle put it, “Not every chicken is created equal.”