The Federal Government’s report focuses on industrial applications, better data foundations and greater collaboration. For CTOs, this comes down to one practical question: Which AI idea is ready to become a project?
Development capacity is limited. AI ideas compete for attention alongside product development, customer requirements and essential technical work. Whether an idea becomes a project should therefore depend not on the appeal of the technology, but on a clearly defined task and a testable objective.
The report “Germany as an AI Nation 2030” states that at least 55 per cent of small and medium-sized enterprises should be using AI by 2030. This is an ambitious target, but it says little about how an individual company should plan its next project. Nevertheless, the report provides a useful framework for development planning. It highlights the questions that technical leadership, management and the relevant business function need to answer together before commissioning a project.
Turn an AI idea into a testable task
Industrial AI plays a central role in the report. The aim is to support models and solutions tailored to individual sectors and to make existing knowledge and data usable for specific applications.
For project preparation, however, this does not answer the crucial question: Where in the business is there a problem that justifies the development effort?
“We want to use AI” is too broad. In a manufacturing quality inspection process, by contrast, a testable question might be whether a particular type of defect can be detected early enough to reduce scrap.
Technical leadership and the development team approach the initiative from different perspectives. Together with management and the relevant business function, the CTO assesses its importance to the company and the level of effort that can reasonably be justified. The development team identifies which assumptions remain untested and what an initial trial could establish.
Sound preparation also includes the possibility of deciding not to proceed. If the trial does not produce a credible prospect of operational value, a well-founded decision to stop is a useful outcome.
No sound project without reliable data
Without suitable data, even a plausible use case remains theoretical. The Federal Government’s report therefore proposes data spaces for the secure exchange of high-quality data, as well as support for companies preparing their data assets. In these data spaces, several organisations can exchange information within a defined technical and legal framework.
What sounds sensible at policy level still needs to be made workable for each individual project. Does the planned application require data that is actually available? Are its quality and timeliness sufficient? Is access permitted, and can the data legally be used for the intended purpose?
A large volume of data does not answer any of these questions automatically.
Even if the necessary data is available for a test, its availability during future operation is not guaranteed. The data must remain accessible, be maintained and have a clearly assigned owner. This work belongs in capacity planning, not in the blind spot between testing and operational use.
The unresolved task determines the choice of partner
Not every open question needs to be resolved internally. The report expressly promotes collaboration between industrial companies, SMEs, AI start-ups and research institutions. It also aims to improve access to solutions, development services, data and computing capacity.
This is not, however, an argument for assembling the largest possible consortium. The appropriate form of collaboration depends on the unresolved task:
- An existing solution needs to be selected and adapted for operational use.
- The company lacks a specific development capability.
- A technical or scientific question needs to be investigated jointly.
These distinctions matter in practice. A specialist supplier may be sufficient to adapt an existing solution. An open research question is more likely to call for collaboration with a university or research institution.
Regardless of the partner, specialist knowledge, data, feedback and decisions about interim results need to remain anchored within the company.
A successful test is not yet operational deployment
A test can be technically convincing and still be far from ready for operational use. This gap is easily overlooked when the trial focuses only on whether the technology works.
Future operation requires more. There needs to be clear business ownership, integration with existing systems and a defined process for dealing with incorrect results. Data protection and information security must also be addressed, along with the question of where human oversight remains necessary.
Not every answer needs to be available before the first test. It should, however, be clear what the trial is intended to investigate and which issues will remain unresolved afterwards.
A positive test result provides a basis for a decision. It is not an automatic approval for operational deployment.
Funding requires a clearly defined initiative
Only once the initiative has been clearly described can potential funding options be assessed meaningfully.
The Federal Government’s report aims to connect existing advisory services, funding programmes and knowledge transfer networks for SMEs more effectively. It also states that an AI voucher for small and medium-sized enterprises is under consideration. Such a voucher would provide funding earmarked for specific AI services.
At present, however, the voucher remains a proposal for review. The report does not specify a confirmed budget or launch date, and it does not constitute a promise of funding.
For existing programmes, the specific initiative is what matters. Funding may help to share development risks. But without a sound project definition and a company contribution in the form of time, data and specialist knowledge, it will not create a viable project.
Statement
The political objective provides an ambitious framework. Whether it leads to a viable initiative within a mid-sized company depends on four factors: the task, the data foundation, clear responsibilities and the unresolved development questions. Only when these have been clarified is there a sound basis for the next step.
From an AI idea to a testable project brief
Ignite supports mid-sized companies in turning AI ideas into testable project briefs, assessing suitable partners and evaluating potential funding routes.
A technical discussion with Ignite can help determine whether the project definition, choice of partner and funding options already fit together.
Book a 20-minute introductory call · Learn more about our advisory approach
Become a partner of Ignite Group today.
In a short call, we check which funding programmes fit your project.
Learn more: Become a partner of Ignite Group today.
