App-based AI Smartly Navigates Away from Failed Deliveries

With the dramatic increase in online shopping since last year, due to the COVID-19 blockchain restriction, more and more items are being delivered instead of being purchased in-store.

Statistically, this also means an increase in missed deliveries, with delivery drivers losing a lot of time and money due to absence.

To address this problem, the JDSC, the University of Tokyo, Sagawa Express, GDBL and the Yokosuka City Council have joined forces to conduct a proof-of-concept test of a delivery diversion system.

In terms of key performance indicators, the application has shown that the total number of missed deliveries during the pilot period, which is expected to run from October to December 2020, has decreased significantly by 20%.

Surprisingly, his project was in the works long before the virus surfaced.

Prototecton: Zero Artificial Conversion Project

The year was 2020. The Japan Data Consortium (JDSC) has introduced a kind of smartphone-based technology amid a rush of COVID-19.

This technology, largely powered by advanced AI, aims to recalculate van routes. As you might have guessed from the title of this article, the most important consideration will be whether the package will arrive at its destination in the right way.

The project has been tentatively named Prototekton, which looks a lot like a dollar version of a super robot from the 1970s.

It is a smartphone application that shows delivery routes and points where a package may not arrive, and calculates the best route to make the delivery in the shortest possible time.

Note that the key set may not be included in the previous statement. You see, artificial intelligence doesn’t just determine whether the recipient is absent or not.

It refers to records of electrical activity in a target’s potential residence and gives it the highest priority among other factors.

We don’t know if looking at months or years of electricity consumption data provides reliable accuracy for AI. If you z. B. you have been very frugal with electricity in recent months, the AI might think you are not really cooling your home and leave you and your delivery items out of it. However, judging by the number of tests conducted so far, it seems to be working well enough to allow for wider, urban experimental implementation.

In terms of development, the very first functional iteration of Prototecton was completed two years before the first. Around September-October 2018, a very small endurance test of the concept was conducted on the campus of the University of Tokyo, which yielded positive results and legitimized its effectiveness.

In addition, the same prototype system was used by Sagawa Express in a computer test in September 2019. The result seemed less promising than the first. However, development of the system has reportedly been expanded to include at least three more groups by the end of October 2019.

First results of field trials in urban areas

The next field test is expected to take place in the last quarter of 2020. The goal was to complete the proposed B route through the city of Yokosuka with the help of five teams. The preparations were announced in July. However, no update has been officially announced until this week’s recent announcement.

The official deadline for final testing of the concept would be October-December 2020.

Using highly updated and integrated data on Yokosuka’s residential areas, the application was made available to drivers of different positions and experience levels (full-time couriers, temporary workers, new hires, etc.) Based solely on user experience, the researchers determined that there was no significant difference in application performance between workers with different experience levels.

In terms of key performance indicators, the application has led to a significant reduction in delivery failure rates of 20% during the pilot period.

It may not seem like much, but when you consider that they were gone the same day, requiring the courier company to be called after, it becomes very significant.

In short, the resulting data image allows the AI system to move from prioritizing previously simpler optimized delivery routes to creating relatively more specific routes to avoid disruptions. Presumably, a sufficiently sophisticated version of this system would ensure that subsequent same-day delivery failures would be the only mention of failed deliveries for the company implementing the system.

Japan supply problems

Japan had already experienced exponential growth in electronic commerce in recent decades, even before the global pandemic hit. As a result, the volume of related activities has increased, but the number of services available online has also continued to grow. This is directly related to the huge increase in carbon dioxide emissions and the need for more engine power to serve an ever-growing industry.

According to the Ministry of Land, Infrastructure, Transport and Tourism, 11.4% of the country’s freight was transported in October 2020 alone. This corresponds to approximately 300,000 deliveries requiring repeated or multiple checks by delivery staff, resulting in a significant loss of time.

Interestingly, the JDSC also shows that for all years in which returns were reported, the average was about 20%. With an average of 25 percent of losses on a covered route and 90,000 delivery drivers nationwide, the organization estimates a net loss of 200 billion yen ($1.8 billion) annually.

As some of you may have guessed, this is an interesting question: How does this technology fit into the rowless technologies we have today? Can the automated navigation systems currently being developed specifically for delivery services be integrated in some way to compensate for the loss of redelivery?

Well, hopefully we’ll have a usable level 5 system soon.


Related Tags:

delivery servicedelivery companyexpress delivery,People also search for,delivery service,delivery company,express delivery