EUVIMEDEUVIMED entdecken
Scoping ReviewOpen Access · CC BY 4.0

On-Device Edge AI for Precision Agriculture: A Systematic Scoping Review

K.Y.B.S. Fernando, K.D. Thamarasee

International Journal of Smart Agriculture · 29.07.2026 · 4 (1) · 46-63

DOI: 10.54536/ijsa.v4i1.7582 ↗

Kurz gefasst

Scoping Review in der Zeitschrift International Journal of Smart Agriculture, online erschienen am 29.07.2026. Verfasst von K.Y.B.S. Fernando und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Edge AI and Tiny Machine Learning (TinyML) have emerged as transformative paradigms for deploying machine learning inference directly on field hardware, addressing the connectivity, latency, and bandwidth constraints that render cloud-dependent systems impractical in real agricultural environments. However, no prior survey systematically maps fully on-device, cloud-independent inference across multiple precision agriculture domains or applies a multi-dimensional evaluation framework to support deployment decision-making. This paper presents a hybrid systematic-scoping review of 41 peer-reviewed and preprint studies published between January 2020 and March 2026, following PRISMA-ScR reporting guidelines. Studies are analysed across four domains (crop disease detection, irrigation and soil monitoring, livestock monitoring, and greenhouse monitoring) using a five-dimensional framework covering technical performance, resource efficiency, economic viability, deployment feasibility, and agricultural impact. Results show that classification accuracy ranges from 92.3% to 99.9% and regression performance reaches R² values of 0.85 to 0.99 across hardware spanning microcontrollers to AI accelerators, yet field validation rates vary considerably across domains and economic viability remains critically underreported, with only 7 of 41 studies disclosing hardware costs. This survey contributes a three-tier hardware taxonomy, the five-dimensional evaluation framework, and a structured analysis of systemic challenges and future research directions to advance edge AI from agricultural prototyping toward scalable real-world deployment.

Zitieren

APA

Fernando, K., & Thamarasee, K. (2026). On-Device Edge AI for Precision Agriculture: A Systematic Scoping Review. International Journal of Smart Agriculture, 4(1), 46-63. https://doi.org/10.54536/ijsa.v4i1.7582

Vancouver

Fernando K, Thamarasee K. On-Device Edge AI for Precision Agriculture: A Systematic Scoping Review. International Journal of Smart Agriculture. 2026;4(1):46-63. doi:10.54536/ijsa.v4i1.7582

BibTeX
@article{Fernando2026,
  author = {Fernando, K.Y.B.S. and Thamarasee, K.D.},
  title = {On-Device Edge AI for Precision Agriculture: A Systematic Scoping Review},
  journal = {International Journal of Smart Agriculture},
  year = {2026},
  volume = {4},
  number = {1},
  pages = {46--63},
  doi = {10.54536/ijsa.v4i1.7582}
}
Kostenlos starten →