Study of applications of Internet of Things and Machine Learning for Smart Agriculture

Abstract

The concept of "smart agriculture" relies on the integration of sensors within an Internet of Things (IoT) network. Machine learning (ML) algorithms are integrated at various levels of the IoT system design to augment its functionality and enhance its capability. This article is a bibliometric review of 42 articles published between 2018 and 2022 using the Web of Science database. The results of the review showed an exponential growth in the use of ML algorithms in IoT systems for different agriculture applications. Additionally, two key research questions are addressed in this article, one being the development of IoT-ML-enabled smart agriculture over the past five years, and the second being the main research gaps for applications of machine learning and IoT in smart agriculture. The article concludes with a discussion of the results and future directions for research in the field.

IoT Solutions with Artificial Intelligence Technologies for Precision Agriculture: Definitions, Applications, Challenges, and Opportunities

“…Table 1 shows review studies that refer to IoT and AI for agriculture. Several studies [61][62][63][64][65][66] acknowledge the complementary relationship between IoT devices and AI algorithms in agriculture. However, these works often lack systematic reviews, potentially overlooking relevant literature and limiting the scope of analysis.…”

“…Through this structured approach, the study aims to contribute valuable insights to the burgeoning field of PA. [24], Ganapathi et al (2023) [6], Abu et al (2022) [25], Shi et al (2019) [26], Kumar et al (2024) [27], Polymeni et al (2023) [28], Shrestha et al (2024) [29], Widianto et al (2023) [30], Fondaj et al (2023) [31], Dewari et al (2023) [32], Zamir et al (2023) [33], Rathi et al (2023) [34], [35], Chataut et al (2023) [36], and Bulut et al (2023) [37]. Cited studies for AI only: Singh et al (2022) [38], Sharma et al (2020) [39], Cravero et [64], Gupta et al (2022) [65], and Baghel et al (2022) [66].…”