Contact at the AI Innovation Center
Holger Kett
AI solutions in productive use
Initial situation
The use of artificial intelligence (AI) offers companies valuable opportunities for competitive advantages. Many companies are therefore interested in testing the suitability of AI solutions for a wide range of application fields and ideas. They are trying to gain initial experience with AI by means of specific pilot projects in which so-called minimal viable products are created for evaluation. In other words, the aim is to develop an initial version of a new product, service or business idea with as little effort as possible. However, a great deal of effort is usually required to initiate and implement the pilots. Processes
do not exist for this. Employees from different areas of the company have to overcome hurdles when working together and building a common understanding. In particular, mutual competencies must be built up with regard to technical process knowledge (on the part of data scientists) and AI expertise (on the part of specialist departments and managers). Cross-divisional and cross-departmental collaboration is necessary in order to drive the topic of AI forward and increasingly integrate it into day-to-day business and internal processes.
AI Innovation Seed - AI solutions in productive use
An initiative in the form of an "AI Innovation Seed" on the topic of "AI solutions in productive use - resilient - compliant - efficient" is therefore being set up as part of the AI Innovation Center.
The initiative addresses challenges and aspects that arise during the sustainable integration of AI into a company and presents solutions, methods and digital tools for this.
There will be the opportunity to discuss specific questions on this topic with experts from the Fraunhofer IAO. Exciting approaches and concrete solutions will be prepared and presented by us in this context. AI Innovation Seeds participants will also have the opportunity to network with other participants, exchange experiences and make contact with other experts on the issues addressed. Furthermore, the specific further development of employees is possible. This forms the basis for a quick and comprehensive introduction to the topic.
The following content-related questions are the focus of the AI Innovation Seeds and the planned future innovation network, but can be adapted and expanded by the participants as required:
- Reference model "AI for productive use": What organizational, ethical and technical requirements do companies need for an AI-oriented IT architecture and its development and operation?
- Organizational aspects: How is the topic of AI organizationally anchored in the company? How can specialist departments be brought together with data scientists? How can a common understanding be created between those involved with different backgrounds? Will centralized or decentralized units or, for example, a data science team be set up? What responsibilities need to be defined? Who takes care of technical issues such as data acquisition and tools? Will AI competencies be built up internally or outsourced externally ("make or buy" decisions)?
- Process organizational aspects: What are the processes in the company for picking up initial AI ideas and incorporating them into an efficient development process? What does the process for operating an AI application ("AI governance") look like and how can this be made permanent? How are AI solutions monitored? What aspects should training include? How can AI solutions be further developed and kept up to date?
- Framework conditions for AI operations: The European Commission published a draft AI regulation in 2021 to support the development of trustworthy AI systems. How do companies deal with legal and ethical guidelines? How are they implemented and what impact do they have on operations and the underlying processes? What does it mean to certify AI solutions? What effort is required, what are the benefits and how are they operated? How does AI support product certification economically?
- Automatisms of efficient AI governance: Which digital tools support the control and monitoring of existing AI applications in a company? Can individual process steps in operations be automated?
The model of "AI in productive use" shown in Figure 1 represents the rough content framework of the initiative. The aim of the AI Innovation Seed is not to cover all levels of the model in full complexity, but rather to introduce the participating companies to the topics of AI organization (1), AI framework (2), operations and governance (3) as well as automation and tools (4) in four events and to discuss the topics and issues relevant to the companies in greater depth.
Each topic (level) is prepared in a half-day workshop and addressed together with the participants. The Fraunhofer IAO offers the following workshops:
- an overview and preparation of the respective topic in concise keynote speeches and interactive workshop formats
- Networking formats at each event for discussion and exchange as well as the opportunity to introduce and discuss your own topics and questions in a larger group
During the first three months, the Fraunhofer IAO also conducts a bilateral interview (in person and preferably with a company tour, or alternatively with a company visit).
online) to take stock of the participating companies. In addition to the objective to be pursued with AI, company-specific technical and organizational issues are also addressed.
The aim of the AI Innovation Seed is to establish an innovation network in 2023 that will be open to other interested companies. The innovation network will
The aim is to jointly address the issues and challenges of "AI solutions in productive use". The companies participating in the AI Innovation Seed can play a key role in shaping the innovation network by incorporating their questions and needs into the "AI in productive use" model. In addition, Fraunhofer IAO conducts a bilateral discussion with all AI Innovation Seed participants to record the initial situation and the specific issues in the company.
The AI Innovation Seed will start in February 2022 with a kick-off workshop to allow participants to get to know each other. This will also include a joint
understanding of the topic was created and initial overarching questions were agreed with all participants. The questions will then be incorporated into the organization and content of the four central events. Participation in the funded AI Innovation Seed "AI for productive use" is free of charge. Companies with headquarters or at least one branch in Baden-Württemberg that can make a relevant contribution to the content of the project are eligible to apply.
The timeline of the AI Innovation Seed:
If you are interested, you can take part in our online information event on November 8, 2021 at 4 pm. Please register by e-mail at niclas.renner@iao.fraunhofer.de and enter the subject "Participation in the AI Innovation Seed - AI solutions for productive use". The information event offers the opportunity to find out more about the content of the AI Innovation Seed and to ask any open questions. We will be happy to receive further suggestions for possible relevant topics.