Redelivery and centralization

To solve the redelivery problem ...

Several issues have been addressed on this site, a redelivery problem.

Currently in the logistics industry, they are taking measures such as setting up a delivery box at the station, home, etc. so that they can receive baggage more reliably.

However, even though it nevertheless arises re-delivery.

Of course, completely eliminating redelivery is not very realistic.

Therefore, we thought about the following solution to slightly reduce the burden on the delivery driver, even when it is necessary to re-deliver without using a delivery box etc etc.

It does not go directly to places where redelivery is needed, and packages that have been re-delivered are consolidated in one place in the delivery area and the customer receives it there It is that.

There is one place, there are many nationwide, convenience stores and others.

Thereby, the delivery driver can save the trouble of delivering the package delivered again at the end of the delivery to each house, it also reduces the burden on the driver for the home delivery, furthermore it does not run a useless distance by redelivery So we can reduce the environmental impact.

By the way, according to the Ministry of Land, Infrastructure, Transport and Tourism, annual carbon dioxide emissions in redelivery nationwide roughly corresponds to annual CO2 absorption of Cryptomeria forest of the same size as 2.5 inside Yamanote line.

About the verification made this time

Verification objective

In this verification, we tried to reproduce the above-mentioned solution actually by themselves.

And, from the reproduced verification, it is to investigate whether the distance becomes shorter if the proportion of the number of redelivery in the visited place is aggregated in one place when roughly approximately.

Method of verification

The verification method of this time is as shown in the above figure ...

① Fix the place to consolidate this package in one place (assuming it is in the middle of the delivery area this time)

➁Set up 20 visiting places mechanically and randomly within the delivery area

At the same time as ➂②, up to 5 redelivered locations are randomly set. (At this point, the driver does not know the existence of redelivering place.)

④Use TSP simulation to get 20 visits to order.

After ➄➃,if the distance when you drop in at the consolidation place set in ➀, the one set of redelivered place set in ➂,the case of two cases · · · · · ·, which is shorter and re- I tried to find out how many deliveries will be effective when delivery to one central location is effective.

➅ After that, do the work of from ten times in total, and obtain the average number of cases where you start to become effective when you drop in at the consolidation place. (As soon as one pattern is over, the site is updated at random and the redelivered area is updated at random as well)


As a result of 10 times of verification, the average number of cases where the time when you stop by the consolidation place starts to become effective is 3.1 cases. In other words, it is when the percentage of redelivered points in the whole visited place is about 15%.


From the results, we found that when the percentage of redelivered sites among all the visited places is about 15% or more, the remedial measures we devised are effective.

However, in reality, there are differences in the number of visits, placement of visiting places, etc. It is impossible to be perfect for values like the result of this time.

Therefore, while re-delivery has been rapidly increasing now, I thought that improvement measures we devised this time can be used with more opportunities.

  • In the case of the current verification condition (the number of visits is 20 in the center of the delivery area), even if the number of redelivered areas is relatively small, the improvement plan devised by us will be effective.
  • In the present age where the number of redelivery has increased rapidly, it turned out that it was a considerably effective measure.