Cooking with artificial intelligencePeter Brown | September 06, 2019
Companies are still discovering how to use artificial intelligence to best suit the needs of consumers and businesses in the future.
It is nearly impossible to find an industry that is not looking to AI for improvements. AI is potentially playing a role in semiconductors, industrial applications, military and defense and everything in-between. Manufacturers hope AI will make developing products and innovation easier.
One of the more interesting emerging uses for AI is in the realm of cooking and food preparation. This new market for AI is focused not just on how to better prepare a meal or make cooking easier, but also on using sensors and AI to make new food concoctions.
“Companies that are developing MEMS sensors are turning to artificial intelligence to identify elements such as smell, taste, consistency and other properties which AI will use to categorize food,” said Luca DeAmbroggi, senior research director for artificial intelligence at IHS Markit. “My view of AI is that it could add value in all those experiences where humans are fully involved with several of their sensors, supporting functions that are perfected with years of experience and trials. This is a great example of the power of machine learning.”
New food innovations
Leveraging AI, meal-enhancement vendor McCormick developed new products for consumers to enjoy at the beginning of 2019. The company collaborated with IBM Research AI for Product Composition to develop better-tasting products as well as new flavor concoctions using sensory science, consumer preference and flavor palettes.
Through this collaboration with IBM, McCormick created new flavors including Tuscan Chicken, Bourbon Pork Tenderloin and New Orleans Sausage. Hamed Faridi, chief science officer at McCormick, said the company is planning on adding more flavor combinations in the future through AI in the future. In fact, Faridi said by 2021, all of McCormick’s 20 labs in 14 countries will be able to use AI to create new insights and share data on new flavor combinations and consumer preferences.
Because flavor is a combination of taste, smell and vision, Faridi said, "It is impossible for a human to reason through all this data." However, he continued, "An AI system can explore the entire space of possible flavors, to find patterns and suggest ingredients that go well together or can be substitutes, as well as their amounts."
In addition to new food creations, McCormick will use AI to help accelerate the product development process that will enable new products to get to market quickly, he added.
Robots in the smart kitchen
How consumers cook is also undergoing an AI-based upheaval. The kitchen is becoming more intelligent with the rise of smart appliances allowing consumers to pre-program cooking schedules or remind users when meals need attention.
Numerous companies are now experimenting with using robots to help prepare meals without any human interaction or help with prep. At this week’s IFA 2019 conference, Samsung is showing off its Bot Chef, a cobot that assists in the kitchen with both food preparation and clean-up. Bot Chef, originally introduced in February, is capable of tasks such as chopping, whisking, pouring and cleaning. The robotic arm has six degrees of freedom and the ability to lift just about any everyday kitchen item.
Samsung said the robot uses both internal and external sensors and AI-based planning algorithms that allow it to interact with humans. The AI and machine-learning skill platform let users program tasks to perform using voice control, physical manipulation and app-based controls. Skills can be downloaded and shared for other people in other kitchens to use. Samsung is targeting the Bot Chef for small and medium businesses looking to save money and accelerate prep.
Meanwhile, Nvidia Corp. opened a new robotics research lab in Seattle in January. Part of the company's research examines how a smart kitchen could be improved with a robotic arm that can perform tasks such as retrieving objects from cabinets, cleaning a dining table or helping cook.
The system uses Nvidia’s GPUs, along with sensors and deep learning technology, to detect and track specific objects. It can also keep track whether a door or drawer is open, or if it needs to open or close them to get access. The GPU processing helps the robot keep track of its environment in real-time using sensors for accurate information to adapt to changes in its environment.
Restaurants implementing AI
However, not everyone wants to cook at home despite having a robot do it for them. According to the National Restaurant Association, restaurant sales are projected to reach $863 billion this year up from $590 billion in 2018. In a recent survey, the organization found that four out of every 10 Americans find restaurants are an essential part of their lifestyle and nine of out 10 enjoy going to restaurants on a regular basis.
With high growth expected from restaurants, it is not surprising that companies are targeting AI for these establishments in order to streamline customer experiences, improve employee efficiency and help address labor shortages during peak times.
Valyant AI has created a restaurant-based AI platform that includes voice assistants that can both listen to customers as well as employees. Unlike Amazon’s Alexa or Google Assistant, the system requires a smaller vocabulary of keywords to take meal orders or answer questions from staff, resulting in better accuracy.
The company has already implemented its AI platform in Denver at a Good Times Burgers & Frozen Custard shop.
Fast-food giant McDonald’s is also experimenting with AI at select drive-thrus in the U.S. The technology can be used to customize the menu or feature promotional items as well as for self-order kiosks, its mobile app and other parts of the business such as the kitchen.
The company said in the future the technology could be used for license plate recognition to identify repeat customers and then showcase a tailored menu based on purchase history.