Vol. 79, No. 3, May 2024

TACHIKAWA Kan

EHARA Koji / MATSUMOTO Yuji

Lately, Japanese industries in the logistics sector  face various issues including compliance with regulations to shorten the long working hours of truck drivers, labor shortages accompanying a declining birthrate and an aging population, and the increasing volume of goods in small lots driven by e-commerce transactions, notwithstanding their role in maintaining a stable supply chain through transportation. This situation underscores the growing need to ensure sufficient labor to handle such heavy workloads.

The Toshiba Group has taken the initiative in constructing cyber-physical systems (CPS), including automated equipment capable of replacing or supporting personnel in physical distribution warehouses, and a warehouse execution system (WES) enabling personnel to collaborate with machines such as robots, under the banner of “best matching between personnel and machines.” We are currently making efforts to promote further automation of physical distribution operations and to create and provide new value to customers and society.

SAKAKIBARA Shizu / KUMEKAWA Kanau / YOSHIDA Takufumi

With the increasing burden on workers in physical distribution warehouses, there is a strong need to introduce automated processes for picking a high volume of goods from shelves and for streamlining work. 

To maximize physical distribution warehouse operational throughput where automated guided vehicles (AGVs) are being used, Toshiba Corporation has developed two optimization technologies for pod transport robot systems capable of automatically transporting dedicated shelves, known as pods, to each work station: (1) pod processing order technology, and (2) AGV plan optimization technology. We have constructed a new pod transport robot system incorporating prototype engines using these technologies at a customer site and have confirmed their effectiveness via actual operations, showing that they can improve shelf picking efficiency by 10%, and can reduce waiting time at shelves by 19% compared with the conventional system.

IWASAKI Toshio / WADA Yusuke / KONDA Kazunobu

In recent years, it has become increasingly important to reduce the burden on workers in the physical distribution domain due to ever more complex business operations accompanying the expansion of e-commerce and workforce shortages. As it is difficult to fully automate picking tasks to handle various types of goods in terms of recognizing and grasping, there is growing demand for labor-saving solutions through the collaboration of robots and workers.

Toshiba Infrastructure Systems & Solutions Corporation has developed an intelligent picking robot system that delivers portability and availability, and is capable of operating in cooperation with workers in response to continuously changing worker allocation in a physical distribution warehouse. Regarding portability, the use of a collaborative robot arm and safety laser sensors eliminate the need for safety fences, reducing the footprint by 92% compared to conventional systems. Regarding availability, the introduction of a recovery processing function, etc. is expected to achieve a business continuity rate of 99.999% in the event of picking work failure. We are now conducting verification tests using a prototype system.

FUJIOKA Kentaro / IWABUCHI Yuta / YOSHIDA Takufumi

Following enforcement of regulations limiting overtime work for truck drivers, Japanese industries in the logistics sector face critical issues, including transportation disruption and labor shortages. To transport the increasing volume of goods swiftly and efficiently, it is necessary to decrease truck wait times at berths through operational collaboration at physical distribution warehouses.

The Toshiba Group has developed technology that makes it possible to allocate the optimal loading berth from among multiple berths and optimize truck loading timing in consideration of warehouse operations. This technology can also optimize complicated situations based on whether there is a need to load goods at multiple buildings, etc. Comprehensive evaluations show that it can decrease wait times between one-fourth and one-eighth of the benchmarking technology. The technology is now available as a function in the LADOCsuite/WES warehouse operation optimization service.

MATSUMURA Atsushi / MATSUO Takuya / WANG Yacheng

With logistics industries facing a critical labor shortage in recent years, the pace of introducing automated equipment in various operation processes at physical distribution warehouses has been accelerating. To improve the efficiency of such processes, it is essential to coordinate automated equipment operations in each process to avoid delays in work schedules, which can lead to interruption of operations, as well as to assigning difficult tasks, which are not suitable for automated equipment, to workers as needed. 

The Toshiba Group has developed a warehouse execution system (WES) to enhance the operational efficiency of automated equipment while balancing tasks by workers via the following: (1) an integrated progress management function for both automated equipment and workers, and (2) a function to optimize task assignments. We have conducted evaluation experiments using actual operational data obtained at physical distribution warehouses and confirmed that the above functions can enhance the delivery throughput.

ONO Soichiro / FUU Shimou / FURUHATA Akio

Various businesses use optical character recognition (OCR) systems, and improving accuracy of extracting information from various business forms, such as semi-structured forms, is increasingly required to deal with labor shortages and to streamline work. The trend of applying the latest artificial intelligence (AI) technologies in OCR systems has accelerated recently, however, the higher development costs for finetuning processes of AI models are serious issues.

Toshiba Digital Solutions Corporation is developing a technology to recognize semi-structured forms by adopting AI language models to its existing AI-OCR service. Our experiments on entity extraction of substrings from the results of recognized strings have confirmed that it achieves extraction accuracy comparable to that of existing methods while saving on advance preparations. In the future, we will consider applying generative AI to allow users to easily customize this technology.

HASEBE Yoshifumi / MORIMOTO Jun

Cables installed in an electrical product play a key role in the transmission of electrical signals and electricity between product constituent components. If a cable connected to moving mechanical parts of the component deteriorates due to repeated bending, it could result in cable disconnection. Because long-term testing is required, it is difficult to confirm whether disconnection occurs during the operating lifetime via prototype testing , so selection of both durable wire material and appropriate cable routing at the cable design phase is essential to improve product development efficiency.

With this in mind, Toshiba Tec Corporation has developed technology capable of predicting the number of bending cycles leading to fracture and/or disconnection of cables in products through the application of a numerical analysis method that takes into consideration cable bending operation stopping time. This technology has been applied to actual product development and is significantly contributing to improving bending durability.