Meituan’s takeaway optimization algorithm: reducing orders may be abnormal and giving riders certain options for taking orders – yqqlm

IT House news on May 10th, today, Meituan said through its official public account that some riders have reported some delivery time issues, including “when there are several orders in hand, the delivery time cannot be arranged”, “merchant can’t be busy.” If I don’t eat for half a day, I don’t have enough time.” and so on. During the delivery process, riders may encounter some small probability events.Meituan internally refers to this type of incident as an abnormal scenethe suddenness and unpredictability of abnormal scenarios will make the originally reasonable delivery time insufficient.

Regarding the response to “abnormal scenarios”, Meituan said it is currently doing two types of things:First, the order angleby predicting and identifying the occurrence of abnormal scenarios, making the estimated delivery time of the order more reasonable and comfortable;The second is the rider’s angleby giving the rider more autonomy to choose, so that the rider can respond flexibly.

IT House learned that Meituan said that for the identified abnormal scenes, itsThere will be two ways of “single stoppage time + system correction”, to give riders more reasonable and ample time for delivery. The former method helps deal with current anomalies, and the latter method helps deal with possible future anomalies. Such improvements cover all aspects of distribution.

In addition, it is not enough to solve the abnormality of the order, what should the rider do when encountering abnormality outside the order? For example, the electric vehicle is broken on the road and cannot continue to be delivered, or the rider takes the wrong meal and needs to return and deliver it again. At present, for such scenarios, Meituan said that it will take into account the user experience and meet some conditions under the premise.Give the rider a certain option to take orders“For example, the rider can set the ‘simultaneous order quantity’, and for another example, when the rider is inconvenient to take new orders, he can initiate a redispatch, transfer the order, and change the order status to ‘busy’ on the App. way to independently manage the current order rhythm.”