The Rise of AI in Restaurants: From the Dining Room to the Kitchen

19 December, 2019

As the digital world becomes ever more pervasive, the food and beverage industry is utilising new technologies such as artificial intelligence (AI) and robotics, in increasingly innovative ways. Within this industry, the restaurant sector in particular is now embracing such a forward-thinking approach and is drawing on these technologies, not only to increase efficiency and productivity and streamline business models, but also to satisfy the recent seismic shift in customer expectations for a great customer experience, which itself has been fuelled by new technologies such as social media and the internet.

There are many ways in which such technology is already, or is set to in the near future, transform the dining experience – the following is a quick look at some of the more interesting ways in which this is happening:

Improving and Personalising the Customer Experience

Restaurants are using AI tools and new technologies to more consistently engage with their guests and create a personalised experience and give diners more of what they want, when they want – be it Millennials searching for a great Instagram photo or time-poor office workers looking to be able to order their lunch quickly and easily.

Self-Serve Kiosks and Touch Screen Ordering

Self-serve kiosks equipped with touchscreen tablets – through which customers place their own orders and make payment by card – are becoming increasingly popular, especially in high-volume quick-service and fast-casual restaurants. In the UK for example, McDonalds, Wendys and Tossed all offer self-service ordering, and more restaurants are sure to follow over the coming years.

This type of system speeds up service by allowing customers to browse the menu and place orders at their leisure and then collect their orders without queuing. Studies have shown that self-ordering leads to higher sales for food outlets as it gives diners the freedom to spend time studying photos of dishes, ingredients and nutritional information, as well as easily request more food or drinks as the meal progresses. As these systems help reduce the repetitive tasks required of staff, such as taking payments and explaining menus, it affords them more time to interact with customers and so provide a better overall customer experience.

Some restaurants, have even started using camera-based facial and biometric recognition technologies as part of their self-service kiosks so they can make menu suggestions and place orders based on customer’s previous eating history. Such tech is already in use at American fast-casual chains Wow Bao and BurgerFi and in some KFC outlets in China, where once diners have opted in to authorise the use of facial recognition, they can order and pay for their favourite dishes in only a few seconds.

A similar system is being trialled in the London cocktail bar Harrild & Sons, to decide who’s next in the queue at the bar. It is using a webcam equipped with face scanners to keep track of new arrivals, and each customer is assigned a number, which is then displayed on a screen telling the bartender who is next. This technology has been developed by DataSparQ and there are future plans for it to be used to allow customers to set up a virtual bar tab without handing over their credit card, and even to help with ID and age checks on customers entering the premises.

Although this type of personal recognition technology undoubtedly improves the customer experience it comes with its own set of problems that include data protection and privacy issues – so if not used correctly has the potential to drive customers away rather than attract them. Despite this, a recent global study by Oracle, a California-based multinational computer technology corporation, found that a third of the sampled restaurant operators think that guest recognition via facial biometrics will be in use within the next five years, while one in two diners agreed that using this would improve their overall experience.

Looking to the future, we’re also likely to see the rise of voice assistant-enabled ordering (as found on home devices such as Amazon Echo or Google Home) in the restaurant sector, which could be integrated into self-serve kiosks or phone ordering systems. Alongside this, wearable technology, such as watches and wristbands that store menu preferences for a particular restaurant or hold personal dietary and allergen requirements, and even virtual reality and drones are other areas in which technology is set to significantly change the way customers and restaurants interact with each other.  

Unsurprisingly, there’s even been a move towards attempting to introduce fully self-service restaurants – where customers don’t come into contact with any staff at all, from placing orders to receiving their food. However the few places that have trialled introducing such systems, such as Eatsa in the USA, have had little success and most have now ceased trading. It would seem that despite customers being more than happy for technology to help improve their eating-out experience, they’re not quite ready for it to remove all personal interaction and effectively turn the restaurant into nothing more than a high tech vending machine with seating.

Reservations and Payment Systems

In addition to improving and streamlining the ordering process, artificial intelligence and associated tech is also doing the same for restaurant reservations. Although there are many apps and websites that will book you a table in a restaurant, there’s a new generation of systems being developed that not only do this but also allow you to order and even pay in advance. One such system is Allset, a smart phone app which operates in the USA, and allows customers to make reservations, place orders for dine-in or take-away food and pay the bill in advance at more than 2,000 restaurants. It even allows users to invite friends, make group reservations and then split the bill at the end.

The company’s main focus is fast-casual restaurants, although it is working with some fine dining restaurants as well. Once a restaurant has received a customer’s order details and payment through the app the food is then scheduled to be ready within minutes of arrival, meaning diners save time and the restaurants efficient service helps maintain customer loyalty. According to the Allset CEO, Stas Matviyenko, the system works by using AI and machine learning, allowing it for example to analyse peak hours and capacity at restaurants to decide how many orders should be sent and at what time. The partner restaurants get paid (and tipped) in advance and know the exact time when a client is scheduled to arrive.

Future plans for the company include launching an updated version of the app that will track customer’s preferred restaurants and food types, and provide personalised food recommendations based on diet and ordering history.

Customer Analysis and Insight

While all these innovations undoubtedly improve and personalise the customer experience they’re also an incredibly useful tool for collecting data that then allows for all-important customer analysis and insight. In fact in many ways this is actually the most useful application of AI and new technology for the restaurant industry as it is allowing operators to make their business models as efficient as possible, both front and back of house.

Restaurateurs can utilise data from self-serve kiosks to predict peak times and customer flows so as to minimise food wastage and maximise staffing, and also to determine customer demographics and preferences so as to be able to tailor meus accordingly or easily cater for diners with particular dietary requests and food allergen issues.

A step on from this are fully-integrated AI systems which can offer even deeper insight. For example, they can be used to predict sales and suggest staffing levels based on factors such as weather forecasts, local events and historical trading patterns; monitor stock levels and place orders; suggest pricing based on cost of ingredients and actual prices charged by businesses nearby; or even recommend discounts or special offers on menu items that use a perishable ingredient that is going out of date or is overstocked. Such capabilities, which will only become more encompassing over the coming years, all work to streamline efficiency and by doing so  free up time for restaurateurs and staff to focus on areas of the business where maybe AI and high tech systems struggle, such as customer relations and good old fashioned hospitality.

Robots replacing Chefs and Other Staff

AI driven automated systems are also starting to find their way into in the kitchen, with robots being developed that assist with preparing, cooking and assembling the food. For example, there are now a few restaurants in California, USA that are utilising these type of robotic systems in their kitchens. Caliburger in Pasadena, has been trialling the use of Flippy, a robot that can flip burgers or fry baskets of food, while pizza delivery company Zume uses AI driven machinery to knead dough, apply tomato sauce and lift the pizzas in and out of the oven. Although at Zume the more complex task of adding the individual toppings still has to be done by hand, the company is in the process of raising capital in order to further develop its robotic technology, with a view to selling it on to other food operators.

Flippy in action

This sort of technology is particularly useful for repetitive and menial tasks, and is therefore once again best suited to fast-casual style restaurants with uncomplicated food and short menus. Its impact on other styles of restaurants is likely to be much smaller and slower, as proper, more complex cooking involves a subtle understanding of ingredients, flavours and textures, and experts have predicted that this is a challenge that may actually be beyond any future AI system’s ability – so we’re probably still a long way off from robots actually completely replacing chefs.

One company that is pushing this technology even further is London based Moley Robotics, which is developing a fully functional AI-driven kitchen. At the heart of this is two advanced robotic arms with hands that are fixed over a stove and then programmed to prepare dishes from raw ingredients. According to the company, they can reproduce the entire function of human hands with the same speed, sensitivity and movement, and pick up and interact with kitchen equipment such as blenders, knives and hobs, as well as handle and prepare ingredients. Interestingly, these robotic arms have been modelled on the actual arm movements of the 2011 UK Masterchef winner, Tim Anderson (and owner of Japanese soulfood restaurant Nanban in Brixton). Although the Moley kitchen is currently only being designed for domestic use there are plans to develop a commercial version and as such offers an insight into how restaurant kitchens of the future might look.

The Moley kitchen of the future

Despite this, robotic cooking systems are very much still in their infancy and have fairly limited functionality, so at present they probably hold more novelty value than practical value. However given the level of research and development being undertaken in this field, there is enormous future potential for the restaurant industry – just so long as diners can be convinced to eat in a restaurant where their food is cooked by a machine.

Restaurant Design

Restaurant design is yet another area that is now being driven by AI and new technologies, whether it be the physical restaurant design or the internal acoustics. In fact this can even extend into designing the overall dining experience, such as working out how best to lay out the dining room with tables or choosing the most appropriate background music. For example Swedish architecture and ‘experience and concept design’ company Livit uses research driven by artificial intelligence to measure and track all kinds of restaurant metrics from what happens when the music gets louder (it inspires diners to purchase more unhealthy foods), customers’ movements, the influence of different plating options for food and even what days require different seating arrangements. This is all then used to inform future design projects and so create restaurants that offer the best possible guest experience, and so also the best possible business potential.

As AI driven systems and robotics become more sophisticated there can be little doubt that they will only become more ubiquitous in our restaurants as operators look to become more efficient, reduce overheads and improve the overall customer experience. Despite this we are probably a long way off the arrival of fully automated high-street restaurants, especially as recent studies have suggested that customers believe that having a personal interaction with staff, and having their food cooked by real humans is still very much integral to what people want and expect as part of a restaurant dining experience. In fact it’s seems likely that with this technology now increasingly commonplace in the restaurant industry, in the future it’s going to be the personalised services, human touches and a proper sense of hospitality that will be where the real value in eating out lies.