A Multidimensional Diagnosis of Smart Farming Technological Capability Across National and Regional Analytical Units: A Potential Basis for Sustainable Agricultural Innovation

Abstract

Smart farming is a convergent technology domain whose upstream scientific and technological capabilities may support sustainable agricultural innovation, but a single aggregate ranking cannot show how those capabilities differ across output types and technology groups. We analyse 28,631 country-assigned Web of Science publications and 64,447 WIPS patent families (2006–2024) across five analytical units—including a fixed EU-27 aggregate—and four technology groups. An integrated diagnostic framework combines three non-substitutable axes: relative rankings of full-time-equivalent (FTE)-normalised publication and patent quantity and of citation impact; technology group configurations of revealed technological advantage and average citation impact; and conditional temporal positions from independently modelled BERTopic–First Appearance Year corpora. Full-period leadership was distributed among the EU-27, China, and the United States; recent joint specialisation–impact positions differed between publication and patent outputs; and first-observation participation centred on different units across the two sources, with narrow temporal leads changing under alternative specifications. Joint interpretation distinguishes robust, divergent, and specification-sensitive patterns to position source- and group-specific upstream capabilities and identify candidates for subsequent outcome research. This role is hypothesis screening, not an evaluation of sustainability performance or policy effects.