What is the Future of Automation? Using Semantic Analysis to Identify Emerging Technologies
Sugat Chaturvedi, Ekaterina Prytkova, Tommaso Ciarli, Önder Nomaler
Identification of emerging digital automation technologies is critical to understanding the changing patterns of work, firm and industry organisation, and labor demand, and thus formulating policies to mitigate the associated risks while harnessing their potential benefits. In this paper, we analyse a large corpus comprising millions of patents and scientific publications from Derwent, PATSTAT, and OpenAlex databases related to automation technologies across a wide range of domains, including but not limited to industrial robots and artificial intelligence. To identify emerging technologies, we propose a methodology which combines machine learning methods with state-of-the-art sentence transformers from the field of computational linguistics. We first identify radically novel patents and publications using a novelty detection algorithm and their semantic off-shoots. We then cluster them into cohesive technology groups based on similarity in their content. We validate these clusters based on obtained labels and observe that citation patterns across patents and publications are heavily dependent on semantic similarity. Finally, we construct aggregate indicators of emergence for these technologies and characterize these based on trends in novelty, bibliometric impact, uncertainty, and growth rates during the past decade. We identify six patterns of technological and scientific development, which provide a better understanding of which digital automation techologies are likely to emerge in the near future, and which have matured. The resulting data set of emerging technologies will be useful to practitioners, policymakers, and researchers interested in the implications of these technologies on labour markets and the society.