The increasing use of AI in research has implications for the epistemological foundations of science and is changing the way scientific knowledge is produced. This also affects our own work as innovation researchers: if scientists do not want to become mere “prompt engineers”, they must focus on the genuinely human core of their work.
On the other hand, AI expands the scope of possible knowledge and can greatly accelerate research. The question remains as to how AI-led and scientist-led research can be balanced in the future. At Fraunhofer ISI, we are intensively and critically examining the issue of embedding AI in our research and are also keeping an eye on the broader implications for science.
The aim of the project was to identify potential new areas of application for a company specializing in industrial cleaning processes and to generate new business ideas. To this end, AI was used to systematically research and analyze current and emerging markets (over the next five years), for example in the automotive industry, where Ecoclean’s existing technologies and expertise could be applied, considering current technological developments and trends.
Among other things, the AI tool AWAIT developed by Fraunhofer ISI was used in this project. AWAIT stands for “Automated Website Analysis for Innovative Technologies”. It is a tool that can be used to accelerate competence analysis for companies and make it data driven.
The aim of the project was, on the one hand, to analyze for the company how AI can be used in strategic forecasting. On the other hand, the project aimed to develop a hybrid, AI-supported scenario process to support the company in strategic decision-making. The findings have been incorporated into the company's innovation process.
To enable innovation with recycled fiber-based plastics, it is first necessary to obtain information about which companies recycle which materials and to what extent they make them available. An important step in establishing a functioning circular economy for fiber-based plastics is therefore to first identify companies and characterize their range of activities.
In the FiberConnect project, Fraunhofer ISI used AI technologies to develop a continuously updated overview of players along the value chain, thereby enabling better networking within the industry and closing material cycles. To this end, an AI-supported, largely automated solution was developed: a web crawler continuously searched the internet for relevant companies, while machine learning models used natural language processing to automatically classify the companies found into predefined categories, such as manufacturers, processors, and recyclers.
The research results were made publicly available on the platform: https://wir-recyceln-fasern.de/en/.
New technologies, materials, and raw materials, as well as new process and logistics knowledge, are among the most important drivers of innovation for companies. But which developments are relevant, and what impact will they have on tomorrow's value creation?
The project “Value Creation Radar: AI-supported forecasting for identifying value creation-relevant signals” developed methods for data-supported and expert-led forecasting. The AI-supported “radar tool” helped to perform semi-automated scanning of data sources to find signals relevant to value creation. The insights gained in the project help companies to set research priorities, adapt product strategies, and better secure long-term decisions.
One shortcoming of national and international studies on knowledge transfer is, on the one hand, the incomplete coverage of participating scientific institutions and, on the other hand, the incomplete coverage of companies involved in the transfer and the various formats of knowledge transfer. This is where the KIWi project comes in: an AI-based analysis of websites of university research groups and companies provides the basis for an improved analysis of transfer activities.
AI methods were also used to improve the quantitative measurability of the transfer. It became apparent that the information base has changed dramatically over the past 15 years, particularly from the perspective of companies, due to the use of the internet and search engines in particular. In some industries, such as biotechnology, this has led to a restructuring of the division of labor in basic research between companies and public research institutions.
The project aims to analyze the opportunities and challenges presented by the convergence of artificial intelligence (especially in the form of deep learning systems) with rapid advances in modern genome analysis and genome editing, and to identify the social and political implications involved.
Fraunhofer Institute for Systems and Innovation Research ISI