For food delivery, machines won’t replace humans any time soon

Robots won’t replace the millions of delivery workers who crisscross Chinese cities every day, carrying food and groceries to consumers in all sorts of weather. Instead, humans and machines will collaborate, making sure tasty meals make it to hungry mouths in a timely manner.

At least this is the kind of scenario that China’s food review and delivery giant Meituan Dianping envisions. “Human-machine collaboration will be the model for autonomous food delivery for a very long period of time,” says Xia Huaxia, the company’s chief scientist, and general manager of the autonomous delivery department.

That’s good news for Meituan’s delivery workers. The company had more than half a million average daily active deliver persons in the fourth quarter of 2017. These people—mainly young men—dart around the streets of the country’s cities and towns, decked out in the company’s signature yellow uniforms. Meituan rival Ele.me had around 3 million registered delivery workers as of late 2018.

Xia cites two examples that illustrate why collaboration will be key. In the case of doorways, sometimes to enter a building or a room, one needs to push. Other times, one must pull the door. For a robot, knowing the difference is a challenge. But for a human delivery worker, it’s a piece of cake.

On the other hand, machines can help human delivery workers by sparing them from those tough duties, such as working the night shift or having to deliver goods in extreme weather conditions, he explains.

The Meituan Autonomous Delivery Platform, or MAD, was launched in July. Trials that begun in March last year, ahead of the launch, explored three main areas around how humans and machines can interact when it comes food delivery.

For delivery in shopping centers, meals were delivered from restaurants inside the mall to delivery workers waiting at assigned collection points outside, saving drivers the time and effort of going into the mall to pick up an order.

Customer trials at office-building compounds entailed delivery in the “opposite” direction—driverless delivery vehicles collected orders from delivery people outside, and brought them to customers inside the building.

The third trial at less-populated areas such as university towns featured a closed-loop operation, with delivery direct from merchant partners to consumers.

In all three situations, MAD incorporates an ordering platform, a dispatch system, and road-network logistics all optimized through big data.

After months of testing, Meituan is planning to gradually roll the system into larger-scale application in 2019, when the delivery robots will work in an area of  with a range of “a few miles,” Xia says, without giving further details about a timeframe.

Happy (working) together

China’s online food delivery market offers “huge opportunities,” says Xia. It surpassed RMB 200 billion ($29 billion) in 2017 and was expected to hit RMB 243 billion in 2018 according to data from research institute iMedia (in Chinese). The report, which was issued early last year, predicted that the number of users ordering food online would reach 355 million by the end of 2018.

“China’s working-age population has plunged by millions after hitting a peak in 2010,” Xia says. Autonomous delivery robots will fill the labor force gap by improving the efficiency of an existing delivery person and also create new jobs for vehicle maintenance, repair and operation, which underlies an upgrading of the workforce, he says.

Meituan’s plan sounds promising so far. But achieving the goal requires joint efforts of various parties involved in the ecosystem, including delivery workers as well as consumers.

Delivery robots don’t represent competition for one 23-year-old man surnamed Zhang, who has been working as a delivery worker for Meituan for six months. He doesn’t think autonomous delivery will come to large-scale application “in the near future.”

At the same time, Chinese consumers, who are accustomed to seamless online meal delivery services, see autonomous delivery as a futuristic concept that is not without potential drawbacks.

Wang Jia, a saleswoman and mother of one 3-year-old girl, came into contact with autonomous delivery while on a trip away. “My daughter was thrilled when we had our meal delivered by an autonomous robot in a hotel,” says Wang. “But having these autonomous vehicles lingering streets is still unimaginable [thing] for me at present.”

The catering industry must catch up with regulation compliance before driverless food delivery can really take off. Although many companies are in the sector, China has yet to come up with regulation covering the industry. While still at an early stage, various parties in the industry are moving towards the formation of a regulation to guide the application of driverless technologies.

In partnership with China Academy of Information and Communication Technology and Beijing Innovation Center for Mobility Intelligence, Meituan has launched a group standard in October 2018 to specify the technical requirements for autonomous vehicles for cargo and food delivery, road cleaning and patrolling (in Chinese).

Last mile, longest mile

Still, fully autonomous delivery remains the Holy Grail of the catering industry, and the connection between automated food delivery and autonomous vehicle is a close one—albeit with important distinctions.

Chinese tech giants in package and food deliveries are the most avid advocates of autonomous driving technology application. In addition to a driverless logistics van that went road test in April 2018, Alibaba’s Cainiao launched its L4-class self-driving logistics vehicle in September of the same year. JD just put its autonomous delivery center in Inner Mongolia into use in last December.

In the case of autonomous meal delivery, the demand for speed and security is obviously lower when compared with that for autonomous driving cars that carry human drivers. Sill, Xia notes that: “Autonomous food delivery shares all the technical problems with autonomous driving.”

Food delivery is a good place for business application of self-driving, Xia argues. It covers a large range of scenarios in different traffic, road and weather conditions. The “application slope” is flat, he said. That means the technology application is easier to get started with lower requirements.

Delivery drones are another popular solution for the last-mile delivery dilemma. Meituan rival Ele.me deepened its involvement into the area but with a different approach by launching commercial delivery drones in Shanghai in May 2018. Chinese e-commerce platform JD won the first national pilot for unmanned aerial vehicle UAV delivery logistics and will be allowed to build tens of thousands of UAV landing platforms or drone pods across China.

Meituan also has been working on its delivery drone for more than one year. It plans to showcase its new product this month.