Addressing the “2025 Digital Cliff” with People and AI - Toshiba Digital Solutions’ DX Strategy (Part 1)

In 2018, the Ministry of Economy, Trade and Industry (METI) warned of the “2025 Digital Cliff.” The report indicated that aging systems could hinder business transformation and growth, worsening IT talent shortages and potentially leading to significant economic losses. These concerns are now becoming a reality. Toshiba Digital Solutions views this crisis as an opportunity for transformation and has been advancing strategic digital transformation (DX) centered on IT modernization. We spoke with Technology Executive KADO Shingo, who has led stepwise modernization, development innovation using generative AI, and engineer training, about specific reforms in development sites, talent development initiatives, and future outlook. In this first part, we explore the “2025 Digital Cliff,” the initiatives to address it, and how generative AI is being applied to software engineering technologies.

(Read Part 2 here)

―As a Technology Executive, could you tell us about your career, including your efforts to improve software development processes and quality at Toshiba Digital Solutions, as well as your role in developing engineers?

I began my career working on projects for clients in the media industry. Specifically, I developed and built systems tailored to a wide range of customer requirements. I later worked on IT modernization initiatives, renewing long‑running systems using the latest technologies, and went on to lead cross‑project support efforts to improve productivity and quality. After leading various solution initiatives, I was appointed Technology Executive for the system integration domain and became head of a division driving software development productivity initiatives. Currently, as a Technology Executive, I oversee software engineering technologies that are essential to our business and work to improve QCD^* in software development.

* QCD: Quality, Cost, and Delivery—an evaluation metric commonly used in manufacturing and project management.

―Could you explain the “2025 Digital Cliff” referenced in the title of this article? What exactly does it refer to, and why has it become such a major issue today?

The term “2025 Digital Cliff” was first introduced in the DX Report published by the Ministry of Economy, Trade and Industry (METI) in 2018.^* It refers to the situation in which legacy systems in Japanese companies have been in operation for many years, with repeated additions and modifications increasing their complexity and turning them into “black boxes.” As a result, maintenance and management costs have increased significantly, requiring substantial human resources. In addition, the lack of transparency in system structures makes it difficult to leverage data across systems. This in turn makes it difficult to invest in new initiatives needed to respond quickly to today’s rapidly changing business and social environments. At the same time, a shrinking workforce is intensifying the shortage of IT talent. Fewer engineers fully understand these systems, creating a situation in which maintenance and modification are becoming increasingly difficult. If this situation is not addressed, companies will struggle to achieve digital transformation (DX), which is essential for enhancing corporate value. As a result, Japan could face economic losses of up to 12 trillion yen annually from 2025 onward.

* DX Report: Overcoming the 2025 Digital Cliff Faced by IT Systems and Promoting Full‑Scale DX, published by the Ministry of Economy, Trade and Industry (METI) in September 2018 (https://www.meti.go.jp/policy/it_policy/dx/20180907_02.pdf (PDF)(2.63MB))

―What specific initiatives is Toshiba undertaking to support DX that delivers value to customers?

We have been working on IT modernization in preparation for the 2025 Digital Cliff. This involves transforming existing systems by applying advanced digital technologies so that they can better respond to business changes, accelerate customers’ operations, and improve operational efficiency. We adopt an approach we call “stepwise modernization,” in which systems gradually evolve into new ones, instead of completely overhauling entire systems at once (Fig. 1). As the term suggests, we leverage portions of existing systems while migrating to new architectures and technologies, starting with the highest‑priority functions. For example, we may first migrate business‑critical operational systems that require rapid change to the cloud, thereby enhancing competitiveness. We then gradually extend modernization to core systems, prioritizing areas that deliver the greatest value to customers. This approach enables strategic system transformation while minimizing migration risks and ensuring steady evolution.

We structure stepwise modernization along two axes—application modernization and infrastructure modernization—and have systematized specific methodologies that customers can select based on their needs. We continuously apply R&D work to modern application development methodologies and integrate software quality management and security technologies—cultivated through our experience in social infrastructure—with IoT, AI, and big data technologies. As a result, we have built a strong track record across a wide range of domains. In addition, to support customer modernization, we offer “Albacore,” a cloud managed service that brings together the Toshiba Group’s experience and expertise to flexibly provide IT platforms, including advanced technologies.

―How does Albacore differ from conventional system operation services?

Albacore is not simply an outsourcing service for system operation and management. It goes beyond system maintenance to drive optimization and automation while enabling continuous improvement. We provide an end‑to‑end service covering IT platform design and deployment through to operations, including operational support. By adopting Albacore, customers can reduce the burden of IT system operations and improve quality, while continuously benefiting from evolving functionalities. Albacore brings together a wide range of Toshiba’s proprietary technologies and expertise, cultivated through our experience in both physical and cyber domains.* For customers, it reduces the burden associated with system updates while enabling them to maintain IT environments that can adapt to change and focus on their core business.

* Albacore is featured in DiGiTAL T‑SOUL Vol. 31 (https://www.global.toshiba/ww/company/digitalsolution/articles/tsoul/31.html).

To support IT modernization, we have established a dedicated organization called the Cloud Center of Excellence (CCoE). The CCoE leads cloud strategy development, standardization, and talent development in an integrated manner. The organization has extensive experience supporting IT modernization across diverse projects and has accumulated significant know‑how by applying successful practices across initiatives. We are confident that we have established a framework to leverage these capabilities and are well positioned to support customers in overcoming the 2025 Digital Cliff.

―Could you explain the advanced software engineering technologies Toshiba has developed to support DX?

DX refers to the use of digital technologies to transform business processes and business models based on data, thereby enhancing competitiveness. Software lies at the core of DX, and fast, high‑quality software development is essential to its implementation. Software engineering technologies—encompassing the methods and frameworks for efficiently developing and delivering high‑quality software—play a vital role in enabling DX. To this end, we have established a proprietary common platform that systematizes the software development expertise we have accumulated over the years. A representative example is “CommonStyle,” a development standard introduced in 2006. It systematizes software development by structuring it into four core elements: “processes,” “development environments,” “application implementation technologies,” and “guides.” By following this development standard, developers can use common procedures and formats to improve productivity while ensuring consistency in software development (https://www.global.toshiba/ww/company/digitalsolution/articles/tsoul/solution/s017.html).

Within the “process” element, we have established an integrated process standard known as “PROSQUARED.” It defines optimal procedures for different development models, such as development on consignment and agile development, and provides templates that can be commonly applied across these processes. It is designed to reduce variation in quality across projects and minimize dependency on individual skills. We have also developed the “CommonStyle Methodology (CSM)” as a design standard within PROSQUARED. CSM consolidates process definitions, templates, and guides to standardize design processes, clearly defining document structures and writing guidelines. This reduces inconsistencies in design documents and enables team members to easily understand and utilize them, thereby mitigating risks associated with knowledge silos. Through this systematization of software development, we can efficiently deliver consistent, high‑quality outcomes even in large‑scale projects (Fig. 2).

―In addition to development standards, Toshiba has recently been leveraging generative AI. How is generative AI being applied to software development?

In recent years, the use of generative AI has rapidly expanded across a wide range of applications. We are also leveraging generative AI in software development to significantly improve productivity and quality while reducing the burden on developers. Our approach to code generation goes beyond simply generating individual code segments. We are developing technologies that enable generative AI to support the entire software development lifecycle—from requirements definition and design to programming—in an integrated manner. For example, in the upstream process, we’re researching the use of generative AI to automatically generate requirement definitions from customer requirement-related information and reviewing design documentation. In downstream processes, we use generative AI to perform step‑by‑step detailed design—such as defining user interfaces and data structures—based on upstream design documents, ultimately working to develop advanced technologies that enable the automatic generation of complete program code.

―I see. So the development standards you mentioned earlier are what enable you to take on these development challenges?

Exactly. By establishing standards such as PROSQUARED and CSM, we have been able to prepare design information in a format that is readily interpretable by AI. This design information is used both to train generative AI models and as input for prompts, enabling integrated support across the entire software development process—from upstream to downstream. As labor shortages become an increasing concern, effective use of generative AI will enable engineers to focus on higher‑value, creative work and deliver greater value to customers. Ensuring quality is a fundamental requirement when using generative AI. We are also exploring ways to apply generative AI in this area while further strengthening quality assurance and carrying forward Toshiba’s long‑established quality management philosophy.

(Read "Addressing the “2025 Digital Cliff” with People and AI - Toshiba Digital Solutions’ DX Strategy (Part 2)" from here)