Logistics DX - AI Applications and Challenges of Advanced Overseas Companies

In Japan, the logistics industry's environment is becoming increasingly challenging each year. Although parcel deliveries have been increasing since 2015, the number of road freight forwarding workers responsible for these deliveries has remained stagnant, leading to a persistent labor shortage. In addition, 77% of the workers are over 40 years old, which means that the industry structure is strongly dependent on middle-aged and older workers, and the shortage of human resources is expected to become even more serious as the population ages. In addition, the Standards for Improvement of Working Hours for Motor Vehicle Drivers (Notification of Improvement Standards) will be revised on April 1, 2024, to impose restrictions on overtime hours and interworking intervals for taxi/hire car drivers, bus drivers, and truck drivers. The so-called "2024 problem," or the situation in which it will no longer be possible to transport goods, is becoming a reality. The "2024 problem" is now a real possibility.

The 2024 Problem
Effective April 1, 2024, the Standards for Improvement of Working Hours for Motor Vehicle Drivers (Improvement Standards Notification) will be revised, and the total number of hours that taxi/hire car drivers, bus drivers, and truck drivers are required to work, the application of an upper limit to overtime hours, and the setting of a lower limit to the amount of continuous rest time per day will be introduced. The working environment for drivers will undergo significant changes. As a result, the total working hours of workers, mainly in the transportation and logistics industry, will decrease, leading to a situation known as the "2024 problem," where goods cannot be transported. In response, the government plans to introduce measures encouraging drop-off deliveries and relaxing delivery time restrictions, with policies being rolled out progressively. In addition, the transportation and logistics industry is facing a major turning point this year, as transportation and logistics providers have successively begun raising delivery fees and charging for some re-deliveries.

This problem is not limited to the year 2024, but we must seek fundamental solutions in the face of a situation that will remain unchanged in the future, where logistics volume will continue to increase steadily while deliverable capacity will continue to decrease. We must also find ways to improve the efficiency of the delivery process through the use of IT, robots, etc. These challenges are not limited to Japan, but are occurring all over the world. In this issue, we will follow the logistics revolution taking place around the world.

Automation of operations required around the world

Reforms in the logistics industry in the U.S. have proceeded from more hard-line means, such as demonstrations and strikes by workers. Multiple demonstrations by Amazon warehouse workers in the US, demanding better working conditions, pay, and the reversal of layoffs, have occurred during major sales events like Prime Day, threatening the continuity of Amazon's services. Similarly, negotiations between unions representing delivery drivers and warehouse workers at United Parcel Service (UPS) in the US and the company regarding working conditions and wages are ongoing, with demonstrations and strikes always a possibility.

From the company's perspective, the more human labor is involved in operations, the more risk there is of service instability and cost increases. This is especially true for food delivery businesses such as Uber Eats and DoorDash, and ride-sharing businesses such as Uber and Lyft, where human labor is at the core of the service.

Initially, these companies secured labor by contracting drivers as sole proprietors for one-off tasks, expecting them to work during their spare time. However, due to the drivers' calls for better working conditions, a bill known as AB 5 in the U.S. state of California in 2019 mandated that Uber, Lyft, and other dispatch service companies treat their drivers as employees (), and in 2021, the Supreme Court in London, England, ruled that Uber and other delivery and dispatch service companies should be required to treat their drivers as employees (). In 2021, the Supreme Court in London ruled that delivery and dispatch platforms such as Uber must treat drivers as employees.

As a result, delivery and dispatch platforms have begun to modify their existing outsourcing contracts and introduce paid vacations, national living wages, and pension schemes for their drivers, not only in the UK but also in Europe at large. Although labor-intensive operations are not exclusive to logistics, automating processes to reduce human involvement is crucial for ensuring stable service delivery and cost efficiency.

Overseas case studies using AI and data analysis

New initiatives are being launched one after another to address these common global challenges. First, we would like to introduce Ocado, a UK-based company that is working to improve the efficiency of warehouse operations. Ocado is an online grocery retailer that sells technology platforms and solutions used in online grocery retailing to the world's leading grocery retailers through its subsidiary, Ocado Solutions.

Features,

• The entire process from order receipt to delivery, including customer order acceptance, inventory management, picking, and delivery scheduling, can be connected through the platform and processed automatically.
• All picking is performed by robots, which process about 10 items per minute. In addition, the warehousing of goods is also done automatically after the inspection is completed.
• The system is equipped with AI and data analysis capabilities to understand customer preferences and purchasing patterns, helping shape effective sales strategies.
• Finally, the system unifies the baskets containing the products and the frames that carry them, and once the baskets are placed on the designated positions on the frames, they can be moved to various delivery trucks simply by pushing the frames, thereby reducing human workload as much as possible.

Additionally, the system automatically shows the location of the delivery basket, the delivery address, and the optimal route, minimizing the time spent on manual searches.

Next, let us introduce UPS's ORION (On-Road Integrated Optimization and Navigation) as a technology that helps improve efficiency during delivery. The system was introduced in 2012 and has been in operation for more than 10 years, and its effectiveness has become more apparent the longer it has been in operation. The effectiveness of the system is more evident the longer it has been in operation. The key to AI and data analysis functions is the algorithm, which is the brain of the system, and the data to be analyzed.

While advanced algorithms are being created one after another, the problem with actually putting them into service is the amount of valid data. UPS has 540,000 employees (as of 2021) and is said to deliver approximately 25 million packages per day. The accumulation of over 10 years of delivery information operating on such a scale on a daily basis provides a significant advantage in the business as a database for routing accuracy.

A few additional details about the business regarding the data to be analyzed are as follows. DHL Germany offers Resilience360, a supply chain risk management system that uses data analytics and AI to provide information that helps companies proactively address risks such as natural disasters, geopolitical events (events, demonstrations, accidents, etc.), and supplier failures. Resilience360 uses data analytics and AI to provide information that helps companies proactively address risks such as natural disasters, geopolitical events (events, demonstrations, accidents, etc.), and supplier failures. Related to the importance of accumulated data in ORION, it is also effective to utilize such external data as a target for analysis by routing algorithms. I have digressed a bit, but next I would like to touch on the question of whether there are any technological innovations that can replace the so-called "last mile," or delivery to the customer itself.

"Last Mile": Where we are today in delivery automation

It was about 10 years ago that Amazon founder Jeff Bezos first announced that individual packages would be delivered from the sky, but unfortunately this has yet to happen. Demonstrations in Cambridge, England, were abruptly halted, and demonstrations in the United States were also halted after 100 attempts. This is said to be due to the hurdles of safety and deregulation, as well as cost efficiency issues. Currently, the cost of drone delivery is estimated by Amazon to be $484 per package, but they have reiterated their goal of reducing this to $63 per package by 2025. Expectations are high for Amazon's drone delivery from the sky.

Meanwhile, in ground delivery, a number of venture companies, sometimes in cooperation with large corporations, are developing autonomous vehicles such as self-driving delivery robots. Levels of automated driving are defined by the Society of Automotive Engineers International (SAE International) based on the degree of human intervention, with Level 4 becoming a reality in the delivery business. Level 4 is becoming a reality in the delivery business.

Automated driving technology standards by SAE International.
These levels are used to track the evolution of automated driving technology and to clarify the roles and relationships between humans and vehicles.
<Level 0: Complete human control>
The vehicle is completely operated by a human.
No automated driving functions are provided.
<Level 1: Driver assistance systems>
The vehicle assists driving in certain functions.
For example, cruise control and lane keeping assistance.
＜Level 2: Partially automated>
Vehicles can perform some driving tasks automatically under certain conditions.
However, human intervention is required and the driver must be alert at all times.
＜Level 3: Conditional Automation>
The vehicle can perform some driving tasks automatically under certain conditions.
However, the driver can intervene only when necessary.
<Level 4: Highly automated>
The vehicle can perform almost all driving tasks automatically.
However, human intervention is required for driving under certain circumstances or in certain environments.
＜Level 5: Fully automated>
The vehicle can perform all driving tasks completely automatically.
No human intervention is required at all, and the driver's presence itself becomes unnecessary.

In 2018, Estonia and US-based startup Starship Technologies started delivering goods by self-driving robots in Milton Keynes, UK, and has since been responsible for delivering various packages mainly within universities in the UK and US, completing over 6 million deliveries so far To date, more than 6 million deliveries have been completed and thousands of Starships (self-driving robots) are in operation every day. The U.S. startup Nuro has already signed partnerships with major grocery retailers such as Walmart, Seven & Eleven, and Domino's, as well as pizza delivery giants, and has begun demonstration tests, An initiative to have self-driving cars actually use public roads to deliver ordered food to customers has been initiated. If the service goes smoothly in Houston, Texas, and Mountain View, California, where businesses, universities, and residences are regrouping, it is expected to be used in an even wider range of areas in the future.

In China, the development of self-driving robots is also progressing, with Neolix, Alibaba Group, and JingDong Group (JD.com) all developing self-driving delivery robots and beginning to use them not only indoors at delivery centers, but also on outdoor driveways inside universities and hospitals. In Japan, startups ZMP and Hakobot have been conducting delivery robot experiments, and Rakuten, Seiyu, and Panasonic have demonstrated and partially commercialized a robot delivery system in Yokosuka, Fujisawa, and Tsukuba Cities, where robots deliver products from local stores to a specified location. Although the service ended in December 2023, future developments are expected.

Advancement of warehouse operations based on human intervention

So far, we have discussed how to reduce human labor in the logistics business. However, there are still cases where human intervention cannot be completely eliminated due to various factors such as the products handled, warehouse configuration, and cost issues. Therefore, I would like to conclude this article by discussing some examples of advanced logistics operations that are useful when human intervention is assumed to be a continuous process.

Smileboard Connect, a warehouse operation enhancement system developed and serviced by Sumitomo Corporation, visualizes the work progress of individual employees at distribution centers and collects, accumulates, and analyzes performance data to improve work efficiency and physical workload. The results of this process can be used to increase work efficiency and improve physically burdensome tasks. What is noteworthy is the way in which the data is utilized. Sumitomo Corporation provides the system to the general public as SaaS, but it is also used at its own distribution warehouses. The system is also being used at the company's own logistics warehouses to identify and reassign workers based on their strengths and weaknesses, which occur only when they are not robots. Shifting workers to tasks they are good at rather than those they are not good at improves their work efficiency, and at the same time, it reduces their mental burden and leads to happiness. In addition, the improvement in efficiency can lead to higher wages for the individual.

Conclusion

The logistics field tends to be long, simple, repetitive, and overworked, making it an industry where job applicants are in short supply and high turnover rates persist. However, the logistics industry is an industry that supports economic activity, and it is no exaggeration to say that the decline of this industry would disrupt people's lives. We hope that a technological revolution will be created in the future that can stably support such an industry and make the people involved in it happy, and we will continue to follow the process of its development. We hope that this article will be of some help to those who are considering business in related fields in developing new ideas and sorting out issues.

(*) Subsequently, companies such as Uber, Lyft, Instacart, and DoorDash launched a massive campaign, investing US$180 million in the passage of Proposition 22 (Prop 22), which would allow app-based workers to be treated as independent contractors. As a result, Proposition 22 passed in California in 2020, allowing companies such as Uber and Lyft to treat drivers as independent contractors under certain conditions.

References.

• 令和４年度 宅配便等取扱個数の調査及び集計方法
• 日本のトラック輸送産業 現状と課題 2023
• 自動車運転者の労働時間等の改善のための基準（改善基準告示） ｜厚生労働省
• As Amazon Prime Day pushes the latest deals, these striking workers want a share of the profits - The Big Issue.
• How likely is a UPS workers strike and how would it affect shipping? | US unions | The Guardian
• Parcel delivery giant UPS avoids first strike in 25 years - BBC News
• Uber UK business model change
• Uber and Lyft 'likely broke law' by classifying drivers as contractors, court rules | California | The Guardian
• Uber and Lyft now can handle drivers as contractors
• Ocado Group
• Resilience360 Annual Risk report highlights top 10 risks for the new decade | Delivered | Global
• DHL RESILIENCE 360
• ‘ORION’ delivers success for UPS | ORMS Today
• UPS To Enhance ORION With Continuous Delivery Route Optimization | About UPS
• Amazon delivery drones: how the sky could be the limit for market dominance
• Starship Technologies
• Press | Nuro
• 自動配送ロボット、中国で「爆発的拡大」の予兆
• 小売配送、自動運転技術の実用化が続々！日米中で（特集：自動運転が巻き起こす小売革命 第4回）
• Rakuten Drone
• 物流業界地位向上に､現場のDX改革で挑む住友商事 | LOGISTICS TODAY