Smarter innovation tracking: What AI-powered web indicators tell us about 3D printing in Germany

What is the study about?

Using ISTARI data, Julian Schwierzy (Technical University of Munich), Robert Dehghan (University of Mannheim, ISTARI), Sebastian Schmidt (Paris Lodron University Salzburg, IT:U Austria, ISTARI), Nils Grashof (Friedrich Schiller University of Jena), Hanna Hottenrott (Technical University of Munich, ZEW) and Michael Woywode (University of Mannheim) are investigating the spread of additive manufacturing (commonly known as 3D printing) in Germany in their article ‘Mapping technology diffusion with AI: A web-based approach for tracking additive manufacturing adoption’. The study combines traditional innovation indicators such as patents and publications with our AI-based web indicator. The focus is on how well the actual use of 3D printing can be measured via company websites – and how this use is related to scientific and patent activity at the regional level.

How can technology adoption be measured using web data?

Instead of relying solely on surveys, patents or statistics, the researchers use our webAI approach: they analyze the websites of over 1.1 million companies in Germany for relevant text passages on the subject of 3D printing. To do this, they apply a deep learning ensemble developed by ISTARI. The model distinguishes between texts that describe the company's actual 3D printing expertise or application and those that only report on the technology in general. Based on the number of relevant text paragraphs, an intensity score is calculated for each website, which describes how important the topic of 3D printing is for each company.

For the traditional innovation perspective, the researchers also take into account regional patent data from the European Patent Office and 3D printing-related publications from SCOPUS at the regional level (NUTS3). The panel is supplemented by regional control variables such as population density, the proportion of knowledge-intensive employment and GDP per capita from the INKAR database.

What does the study reveal about 3D printing in Germany?

The results make it clear that additive manufacturing is still a niche technology in Germany, but it is growing rapidly. The web indicator identifies 4,131 3D printing-active companies for 2022, and 8,273 for 2023 – meaning that the share rose from 0.37% to 0.74% of all companies with an active website. Geographically, the adoption hotspots are mainly concentrated in central, southern, and southwestern Germany, including Baden-Württemberg, Bavaria, and parts of Thuringia. Examples of particularly strong regions are Tuttlingen, the Enzkreis district, and Jena, where traditional strengths in optics, automotive supply, and medical technology serve as catalysts for 3D printing.

ISTARI's web-based indicator shows 3D printing activity in more regions than patents and publications alone would suggest. This indicates that the use, services, and trade surrounding 3D printing go beyond traditional R&D metrics. Web data can also be used to identify service providers, users, and retailers who do not file patents or publish scientific papers but nevertheless contribute significantly to diffusion.

Source: Schwierzy et al. (2026)

What are the implications for policy and innovation promotion?

Several clear recommendations for policy and innovation practice can be derived from the study:

• Use web-based indicators to supplement established STI metrics: Traditional surveys, patents, and publications remain important, but they only reflect part of the reality. AI-supported web analytics can make the actual use of technologies more visible, especially in rapidly growing, interdisciplinary fields and technologies.
• Make regional innovation policy more data-driven: Since the web indicator reflects 3D printing adoption at the company level, targeted measures can be developed in aggregate—such as cluster-oriented funding in existing hotspots or specific programs to activate previously underserved regions.
• Institutionalize diffusion monitoring: Our approach allows the effects of funding programs on actual technology adoption to be observed in real time, rather than only in the long term via patents or infrequent surveys. Especially in the case of key technologies such as 3D printing or AI, continuous monitoring is crucial for decision-makers to be able to react early to undesirable developments or opportunities.

Web indicators as a reliable approach to technology monitoring

Overall, the study shows that AI-based web analytics are not a substitute, but rather an effective additional tool for making technology adoption visible in a granular, up-to-date, and practical way. Patents primarily capture the inventive front, publications the scientific basis, while web indicators shed light on the transition to use, services, and market applications. Consequently, our approach can be applied not only to 3D printing, but to a whole range of key technologies.

Reference: Schwierzy, J., Dehghan, R., Schmidt, S., Grashof, N., Hottenrott, H., & Woywode, M. (2026). Mapping technology diffusion with AI: A web-based approach for tracking additive manufacturing adoption. International Journal of Information Management Data Insights, 6(1), 100387.

Link zur Studie: https://www.sciencedirect.com/science/article/pii/S2667096825000680