Innovation in the tool market: Why is it so difficult?

In my presentation at this year's ReConf, I will address a paradox that frustrates many people in product development: Why is there so little innovation in the tooling market? And can AI revitalize the tool market? After all, many of the tools we still use today date back to the last century: Classical DOORS (1991), Simulink (1990), AutoCAD (1982), MagicDraw/Cameo (1998).

10% ReConf discount: Save with this code 10%: REC25_SPRE_8103 (Registration)

This article is less about AI (as in the ReConf talk) and more about the question of innovation in the tool ecosystem in general, with AI being just one of many drivers:

• Case Study: Requirements Management - Market stagnates despite new web tools - old tools remain in use.
• Comparison: Software IDEs - Software tools developed rapidly and often free of charge.
• Differences: culture, value creation, business model - Innovation slowed down by closed culture, compliance focus and indirect value creation.
• Is more innovation possible in product development? - Cautious yes, role models and competitive pressure create momentum.
• Innovation in product development - The drivers are agility, AI, SysML v2, API-First and web technologies.
• What's missing: Ecosystems - There is a lack of open components, as in the software world.

Case Study: Requirements Management

The market for requirements management tools (RE tools) is a good example of what is happening on the market. In the 1990s, a dozen tools appeared on the market, some of which still exist (e.g. DOORS, CORE, Visure or Cradle). These were all desktop applications, although some have since been given a web interface.

In the late 2000s, there was another wave of web-based tools, such as Polarion, Jama Software, DOORS Next and Codebeamer. These tools were all web-based, had better usability, but were otherwise very similar to the products from the 1990s.

Recently, things finally seem to be moving again, driven in particular by AI and modeling. Candidates from this domain include trace.space, Valispace and Flow Engineering. More on this in a moment.

The majority of tools are rather expensive, with annual license costs in the high four-digit to mid five-digit range for a team of ten employees.

But we must not forget: Word, Excel and PowerPoint are still used for requirements management. Software development tools such as Confluence, Azure DevOps or wikis are also often used. This definitely works for simple or small projects.

Comparison: Software IDEs

The contrast in software development to requirements management could not be more drastic: In the 1990s, expensive tools such as Turbo Pascal/Turbo C or Symantec Cafe were widely used, the alternative being text editors such as Emacs or vi.

This changed drastically in 2001 when IBM surprised the software world with the open source and free Eclipse. This initially made things difficult for commercial providers, who were nevertheless able to assert themselves with an outstanding offering, such as JetBrains (IntelliJ).

Since then, things have progressed rapidly: in addition to IDEs, more and more tools have been added to the ecosystem for software development, many of which are also free and partly open source: code repositories (Subversion, git), CI/CD infrastructure (Jenkins, Maven). In addition to tools, platforms and services were also added, such as gitHub, Jira and many more. Much of what was offered to users, often free of charge, was highly innovative. As a final example, I recently posted here Cursor AI presented.

Differences: culture, value creation, business model

These differences can be explained by three aspects. I will stick to the two examples of RE tools and IDEs:

• Culture: Open vs. secret: Software costs virtually nothing to reproduce once it has been created. This is one of the reasons for the strong culture of openness in software development. Product development is hardly possible without investment, sometimes large ones (production machines, materials, etc.). In addition, the fear of IP loss is often high.
• Drivers: Compliance vs curiosity: The acquisition of RE tools is almost always driven by compliance, which makes costs less important. In contrast, the driver for IDEs is the desire for more efficiency, but often also curiosity. It is precisely this curiosity that providers exploit with freemium models.
• Value creation: Direct & immediate vs. indirect & late: With an IDE, the user is one step away from a functioning product (the compiled code is executable). In contrast, there are many intermediate steps between the RE tool and the finished product. A lot of time also passes between working on the requirements and the first prototype. As a result, the perceived value for value creation is less pronounced.
• Market size: Factor 2 difference: According to Data Bridge the market for requirements management tools will be worth around USD 1.5 billion in 2025. The market for IDEs is according to Verified Market Research almost twice as large.

Is more innovation possible in the tool market for products?

Two questions arise with regard to innovation in product development. Will we see more innovation in the near future despite these differences? And will these differences perhaps disappear?

My answer is a qualified yes to both questions. Qualified, because at the moment it is still unclear whether Europe will act quickly enough.

Hardware and electronics are slowly becoming like software

3D printing has already greatly transformed production. For little money, anyone can afford a 3D printer that delivers acceptable results, at least for many prototypes and MVPs. For higher quality, we can order parts online from a wide variety of materials, which arrive in the post two days later.

There are also many suppliers of electronics who produce and assemble circuit boards for little money. Inexpensive single-board computers can be used for prototypes.

This does not bring us to zero-euro costs for the "production" of software, but to costs that are also affordable for a hobbyist. At least PTC has copied a software development strategy with the acquisition of OnShape (CAD/3D printing): The web-based tool is free for non-private projects.

Innovation despite differences

In my opinion, the differences are only shrinking slowly. Nevertheless, I cautiously hope that we will see more innovation in the near future. The reasons:

• Others are leading the way: SpaceX has worked with self-landing boosters The company continues to develop the Raptor engines on a 2-day cycle. Copying innovation is far easier than trying something that has not yet been proven to work.
• Competitive pressure: China-Speed is currently a buzzword that expresses the fact that speed is essential for survival. However, traditional approaches are not enough. Product companies are therefore increasingly looking for new, innovative solutions.

Innovation in product development

I have already reported here several times about interesting start-ups that have the potential to transform product development. These fall into two categories: Innovative and open.

The drivers that I see for innovation in the tools market are (still) agility, AI, SysML v2, API-First and web technologies.

What all the products presented here have in common: They are all web-based.

Innovation in the tool market through new ideas

• Spicy: MBSE, but user-friendly and without SysML
• Flow Engineering: Seamless model integration
• Valispace: Parametric linking of components
• Raiqon: Accelerate with AI without having to rebuild tools and teams
• Trace.space: Web-based requirements with AI

Innovation in the tool market through open platforms

• SysON: Open, web-based SysML v2 platform
• OnShape: CAD for open source projects
• SimScale: Simulation in the browser

What's missing: Ecosystems!

But what has really accelerated software development is the huge ecosystem of easy-to-use components.

I don't see anything really comparable in product development. Only:

• Standards for loss-free data exchange have now been established for the various areas (ReqIF for requirements, STEP for CAD, etc.).
• Norms and standards (ISO, DIN, etc.) play a major role for reuse; usability is also slowly improving here (e.g. DIN now offers standards in ReqIF format)

But from a living ecosystem of system components, as with npm (node package manager) or pip (Python) we are miles away.

My ReConf presentation: Resilient wishes from the field

Finally, an invitation to meet me at my talk at the ReConf on May 7 at 11:45. In this lecture I try to anchor the AI hype a little in reality. I had already dealt with AI in systems engineering before the age of "before ChatGPT" and collected hard facts in over 100 stakeholder discussions. Since then, I have conducted further customer surveys and also participated in many public and non-public debates.

If I had asked the users, they would have said: A faster DOORS

Michael Jastram, Lecture ReConf 2025

I would be delighted to meet you in Munich!