Where Is the Automation Discourse Heading? Exploring Evolution and Trends of RPA Research Through Topic Modeling
As a discipline that has surpassed its initial hype, Robotic Process Automation (RPA) is now a mature technology driving significant business transformation, which has led to an exponential growth in academic literature. At this juncture, traditional systematic reviews, with their limited scope, struggle to provide the necessary holistic overview. This study addresses this gap by creating a comprehensive, data-driven ”science map” of the entire RPA research landscape. Leveraging an unsupervised machine learning approach, Latent Dirichlet Allocation (LDA), we analyzed a global corpus of over 3,200 scientific abstracts to provide a much-needed perspective on the evolution, intellectual structure, and emerging trends that are shaping the discipline’s future. Our analysis identified 100 distinct research topics, revealing a clear maturation trajectory. We identify Lifecycle Management as the methodological pillar of the discipline, while Healthcare and Auditing stand out as its most prominent application sectors, with the latter demonstrating exceptionally high and sustained scholarly impact. Ultimately, this study provides a cartography of the RPA field, acting as a compass for navigating its complex terrain at a moment of change. We highlight the trend towards the integration of AI and Machine Learning, which is propelling the discipline from simple task automation towards a new, more AI-driven paradigm of “agentic process automation”. This map offers valuable insights for academics seeking research opportunities and for practitioners aiming to understand the strategic trajectory of automation.
Institutions
- Technical University of Liberec (CZ)
- University of Žilina (SK)
- Instituto de Recursos Naturales y Agrobiología de Sevilla (ES)
Publication Details
- Journal
- Array
- Published
- 2026-03-01
- DOI
- https://doi.org/10.1016/j.array.2026.100777
- Primary Topic
- Robotic Process Automation Applications
- Type
- article
- Field-Weighted Citation Impact
- 0.00