Shulou(Shulou.com)06/02 Report--

In the past two years, the AI of the catering industry is in full swing, but there are thousands of ways to Rome.

Unlike many restaurants (such as Pizza Hut and KFC), which are keen to join forces with technology companies for win-win cooperation, McDonald's always wants to make big news, and the method is very simple, just one word-buy!

It paid $5 million for a 10 per cent stake in Plexure in early 2019, followed by Dynamic Yield, an Israeli algorithm company, for more than $300m, and recently completed the acquisition of voice technology company Apprente.

In addition, McDonald's has set up a new team called McDTechLabs to recruit engineers, data scientists and so on in Silicon Valley. An air that wants to compete with FLAG (the four tech giants in Silicon Valley).

We know that there are still many delicate gaps between "buying AI" and "doing AI". Is McDonald's good at art and bold, or is it something you don't want to drink?

McDonald's AI renovation Plan

Even in a remote town, you can feel the public enthusiasm for AI. The most typical is that many friends in the catering industry began to wonder, "if you want to make the store look high-tech, can you add two AI chefs and robot waiters to attract customers?"

No, I don't think, at least McDonald's thinks, maybe. Not so good.

Obviously, McDonald's does not intend to create the future with third-party technical services partners like its peers. It chose to go it alone.

You and I have no chance, it all depends on my money. What if there is no AI gene? Buy it! What if there is no technical team? Keep buying!

First of all, in terms of improving the efficiency of the restaurant, McDonald's did not choose those cool-looking "cool techs", such as letting robots bring you tea, water, singing, dancing and shouting warmth, but focused on the optimization of the internal system.

The core competence of Plexure, the artificial intelligence company that was first acquired, is to help restaurants build an app to process orders quickly. In the future, the plan will also be fully connected with buffet machines and mobile phones to reduce customer waiting time as a whole.

Rely on AI to enhance user experience and efficiency, but also further reflected in a variety of interactive details and strategic layout.

We know that McDonald's has a large number of Delais restaurants, that is, "Drive-thru" driveway restaurants. After the introduction of Apprente's artificial intelligence voice technology, customers can place orders through voice at the entrance, and even complete a series of processes such as ordering, payment and picking up directly without the need for a manual waiter. This helps McDonald's to provide services in more remote areas, reducing the waiting time for consumers.

In addition, the electronic transformation of stores has also allowed McDonald's to accumulate huge amounts of data, which has been coruscated with commercial value with the addition of dynamic yield, a machine learning company.

The company's "decision logic technology" (decision logic technology), which automatically adjusts menus to changes in external factors such as time of day, weather and traffic, in order to use personalized recommendations to increase customer unit prices, is now available in hundreds of stores across the United States.

For example, based on previous purchases, it is recommended that you buy a large coke, or when the weather gets cold, you may need to bring a potato cake with your hamburger. This combination has proved to be very popular on Reddit.

Different ways lead to the same goal: McDonald's allies and bad neighbors

Of course, McDonald's is not alone.

Earlier last year, sandwich chain Subway Subway also bought Avanti Commerce, a Vancouver startup, and this small team of 20 people will become a digital army of about 150 people within a few years to help Subway build large-scale consumer technology.

Not long ago, for example, Subway launched the Facebook Messenger chat robot, which allows people to order food directly from 26500 locations in the United States through Facebook, and then use the built-in MasterPass system to pay for orders. This service is likely to cover 68% of American adults (total users of Facebook), bringing possible growth to Subway.

McDonald's competitors, American restaurant chains such as Sonic and Good Time Burgers, are also testing AI voice systems similar to McDonald's. It is said that the related technology can reduce the average waiting time by 7 seconds, and 80% of customers are satisfied. But the difference is that they don't choose to buy the technical team directly.

In fact, catering technology stocks have been heating up in the capital market, with 80 financing for catering technology start-ups in the United States in the first half of 2017 alone, totaling about $594 million, including star investors such as Sequoia Capital and JPMorgan Chase. Technology companies have contracted almost every aspect of restaurant renovation, from selling software to payment systems, from staff training to food warehousing.

Momentum Machines, which designed the hamburger robot, raised $18.4 million from Google Ventures (Google Ventures) and Khosla Ventures Ventures.

In addition to capital, this model is also more easily accepted by most catering brands. KFC's intelligent ordering is accomplished through Baidu's face recognition technology, and Haidilao's food-passing robot is also realized in cooperation with hardware manufacturers. The robot restaurant "Robot.HE" of Box Horse even claims to sink the solution to more small and medium-sized restaurants.

Obviously, most brands prefer to achieve AI transformation by purchasing technology or cooperation.

At this point, "Restaurant AI" seems to have entered a chaotic state. AI is the destination, but there are many ways to get there. Since it is possible to use AI to buy companies, is it worth replicating the acquisition path that McDonald's has taken?

To buy or not to buy is a question.

The answer may not be so.

At present, it seems that McDonald's, subway and other catering groups have introduced technical teams, which are characterized by a large number of chain stores, a unified supply chain and production standards, and the introduction of AI has a bonus effect. This enables it to export quickly after acquiring technology / products, and reduce the cost of technology on a large scale.

Ed Rensi, former chief executive of McDonald's USA, once told the media that "it's cheaper to buy a $35000 robotic arm than to hire an inefficient potato chip worker who earns $15 an hour." Now that McDonald's has more than 30, 000 fast-food restaurants around the world and serves about 9 million pounds of French fries a day, the amount of money AI can save is clearly worth betting on. Visually, it's time for McDonald's to buy robotic arm company.

In addition, the acquisition of technology companies will also allow McDonald's to tell new and more attractive stories in the capital markets.

Data show that McDonald's investment in technology is translating into the highest increase in years, and McDonald's president and CEO SteveEasterbrook also said that to build their own "technology infrastructure and digital capabilities", whether this ability will be released will also give McDonald's more room for imagination, which is a potential benefit that cannot be realized simply by revamping the overall system.

(McDonald's US stocks performance as of 2019.9.20)

So we will find that before launching AI, the relationship with technology companies is not the most urgent. The first thing the restaurant needs to solve are these problems:

What tasks are most likely to be automated by 1.AI, robots, etc.?

Of course, the premise is that too much cost is not transferred to the customer. For example, getting a robot to serve food is not only difficult to move flexibly in a complex environment, but also requires humans to cooperate with it, which does not bring additional efficiency benefits.

Is the input-output ratio of 2.AI reengineering industrial process reasonable?

Many enterprises have announced that they will use drones and unmanned vehicles to deliver takeout, leaving the delivery staff out of work. But the reality is that it is easier and more effective to hire a few people for remote control than the long-term L5 self-driving technology. For example, the "unmanned car obstacle avoidance" that made Google Apple Tesla bald, startup Kiwi Campus used workers as far away as Colombia to solve the problem. The company's unmanned dining car, Kiwibot, is an online takeout celebrity active in American colleges and universities. However, due to too much entry into the artificial world, it has recently been faced with the public opinion crisis of "pseudo-intelligence" and "fraud hype". The subtlety between artificial and intelligent is really difficult to grasp.

3. How to strike a balance between private data and system intelligence?

Data is the core competitiveness of AI restaurant on arrival. The more accurate the operator's data on customers' age, income level and consumption frequency, the more conducive to the optimization of restaurant products and systems. At the same time, it also means that more personal privacy is placed on the dangerous digital balance.

From this point of view, perhaps McDonald's technology exploration is not somewhat representative, through the acquisition of algorithm companies, you can control the use and training of data in your own hands, which is obviously easier for consumers to let their guard down and accept new technologies.

Generally speaking, McDonald's "AI strategy" also follows three basic principles: reliable-in line with the current technical capabilities and industrial reality; available-mature data base and clear business prospects; controllable-public acceptance and fault tolerance is high enough.

This may indicate that AI is not a smooth road to heaven. Under its bright dome, some people hold the power of money, some people hold the shield of the brave, but who will write the end of the story, everything is still unknown.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.