A food delivery person of Chinese food-delivery and online services provider Ele.me Photo: IC

A viral report by a self-media writer which suggested that delivery staff in China are being further squeezed in terms of their delivery time by the takeaway platforms supported by the AI technologies and database raised concerns among the public who called on the delivery platforms to ensure a more humane management while using AI to maximize income.

The report, which is based on a survey of dozens of delivery persons and industry analysts, found that the two big Chinese food-delivery and online services providers Meituan Dianping and Ele.me have "mysteriously" shortened the delivery time in the past several years based on the AI system, making it increasingly difficult for deliver persons to do the job on time. The report received more than 100,000 views and numerous comments with many echoing concerns.

Ele.me officially announced on Wednesday via its WeChat account that they will launch a new mechanism that will enable the platform not only to provide more delivery time for delivery persons but also ensure orders are delivered on time.

"Systems are programmed and people can be flexible. Ele.me hopes to do better while ensuring that the orders are delivered on time," said the company in the statement, noting that they will launch a new function that allows customers to choose if they are willing to wait a few minutes longer for an order and in return they will receive some coupons from the platform.

The delivery persons with a good service record will not have to take responsibility if they occasionally fail to deliver orders on time, said the statement.

A delivery person of Meituan Dianping surnamed Zhang who has been working for this platform for three years told the Global Times on Wednesday that he also noticed that the delivery time was being shortened by 10 minutes or more for the same distance. 

"When I first started this job, I was given at least 40 minutes or even 50 minutes of delivery time on a single order and now we are generally given 30 minutes for delivery," said Zhang, noting that some restaurants can't even cook a dish in 30 minutes and the result is that he will be punished for late delivery.

Data analysis platform Hummingbird showed that riders are basically on call 24 hours a day, with an average daily delivery of 48 orders and a journey of nearly 150 kilometers. 

The platform normally calculates and automatically generates the delivery time according to the delivery distance, peak time and the amount of the order before distributing it to the delivery persons.

Meituan Dianping also issued an official statement on its WeChat account on Wednesday promising to make changes to ensure delivery staff are better taken care of, which include a scheduling system that offers them an 8-minute window of flexibility while providing customers with on-time delivery services.

In order to ensure the safety of delivery staff, Meituan Dianping said that it will strengthen the delivery security technical team, focusing on how technology and algorithms can guarantee security with the launch of smart helmets for delivery staff.

However, what they cannot control is the time restaurants need to prepare for the dishes which is also included in the delivery time.

"Our pay is deducted for late delivery by a few cents for each order which normally means several hundred yuan of deduction in a month," said Shu, another deliveryman, who chose not to disclose his company's name, noting that they have to sometimes violate the traffic rules for punctuality.

According to a report by Meituan Dianping released in August, the number of the platform's annual active users rose 8.2 percent year-on-year to 460 million. In addition, in the first half of 2020, the number of delivery persons reached 2.95 million. Among them, up to 1.38 million were new riders. 

Liu Dingding, a Beijing-based independent Internet analyst, told the Global Times that the reason why delivery platforms can serve such a large number of customers at the same time is that it relies on a huge scheduling system supported by AI technology to connect the track of riders according to the analysis of historical and real-time data.

"However, as the market economy pursues efficiency, how to establish a more humane and comprehensive system is a problem that related companies have to think about in the future," said Liu.

It is inevitable that problems will arise in the process of the development of delivery platforms, but while looking at the problems we also need to think of their social responsibilities as employers and market needs, said Liu.

How to effectively utilize AI technology and big data to maximize business revenue and at the same time ensure more humane management would be a common issue that the service providers need to weigh more in the future, experts said.