Consulting on digitization and AI in development

Harmonizing digital strategy and information architecture

Applications based on digitalization, connectivity and AI are present in many products: from digital assistants and smart home solutions to support functions in software packages or on smartphones.

In production, digital twins and the IIoT have also been demonstrating their enormous potential for increasing efficiency and quality for some time now. What is already commonplace here, however, has often not yet arrived in R&D.

Development performance: digitalization pioneer and laggard at the same time

Why is this the case? R&D teams have been relying on digital solutions and systems for many years. Large drawing boards have long since disappeared from offices and employees usually work in highly integrated development environments. However, the challenge is the state-of-the-art digitalization of development processes. Here we often find digital islands, between which lies an ocean of paper and data conversions.

Frequently, the integrated, direct exchange of data is not directly possible. Whether this is data on the product or the development process, it is stored in different systems, it is redundant and often diverges. Miscommunication, a lack of transparency and expensive development errors are the result. At the same time, more and more companies want to use AI in an R&D context. This provides new challenges and changes to working methods and infrastructure.

Optimization: smart digitalization and AI as performance boosters in R&D

The main effects of properly implemented and lived digitalization in R&D are

• Greater transparency in the development process.
• Lower error rates due to data inconsistencies.
• Greater opportunities to optimize processes.
• Solid database as a basis for decision-making.

These effects alone lead to lower throughput times and R&D costs – and therefore to a generally higher development performance. The use of AI solutions is accompanied by further performance-enhancing effects. This can already be seen in examples such as

• AI use to optimize mechanical designs.
• AI analysis of log and usage data for better user understanding.
• AI-supported coding for faster software development.
• Automated response to support requests.

Information architecture: digitalization and AI require a solid foundation

The use of digital tools and AI means much more than the next purchase of licenses. This approach only leads to another island and the expansion of existing complexity in the digital R&D landscape. In order to successfully leverage the benefits of digitalization in development, companies need in particular:

• Cross-divisional consistent information models and architectures, from users in business to development to production and supply chain.
• Coordinated data and IT structures as well as data consistency between tools and systems.
• Models and consistent sources of comprehensive, consistent and high-quality data on products and development processes.
• Adapted processes and organizational forms that interact ideally with the digital solutions.
• A holistic, entrepreneurial understanding within the organization.
• New knowledge and an understanding of how to deal with AI and other new tools.

Development efficiency: mastering the establishment of new systems

A major obstacle to the introduction and further development of such systems lies in the creative nature of R&D. Many developers see clear advantages and benefits in specific development systems that support them. The same does not necessarily apply to systems for managing the processes themselves. These often appear to be of little value to developers, and appropriate persuasion and change management measures are necessary during implementation to reap sustainable benefits.

Managers often do not see themselves in a position to obtain a picture of development progress or development performance within a very short time. At the same time, many different domains and specializations are represented in an R&D organization, each with their own specific requirements – and each preferring their own specific solutions. Clear strategies and information models and the corresponding work with the developers are required – in order to achieve a meaningful integration of data and system that is acceptable to those involved and represents added value.

R&D tool landscape: a fool with a tool is still a fool

It is immensely important to know the limits of digital tools and to make them clear to all employees. This is because the incorrect use of tools or blind trust in them poses major risks, which in the worst case can result in liability cases.

Far too often, developers think that the introduction of a tool alone will solve many of the difficulties they experience. Too rarely do they see that these problems arise from inadequate processes, inappropriate organization, and poor collaboration, which a particular tool alone does not solve.

AI as a risk factor

There is too much trust in the performance of AI technologies and too much enthusiasm for their capabilities. So, results are adopted without checking and critical errors made by these tools are incorporated directly into the products. It is frequently not clear to employees what responsibility they have to take on here. And too often, managers do not realize that the handling of the tools is more important than the tools themselves.

This is because the implications of the latest technologies for the respective area of application must be examined and made transparent to everyone. Just as many organizations are now taking action against the misuse of their intellectual property by AI operators, they do not want to "accidentally" make their trade secrets and intellectual property available to the general public by using an AI tool.

Shaping the digital R&D future with EFESO

The management consultancy EFESO offers you a range of services to shape the digitalization of your R&D organization and thus increase your development performance. The focus is on the following fields of action: