Artificial intelligence (AI) combined with robotics is changing how agriculture has always been practiced to a more efficient, precise, and sustainable one. The review critically discusses the latest achievements in AI-driven agricultural robotics, including vision-based navigation systems, deep learning-based crop surveillance, and autonomous robots based on hybrid control architectures. It analyses fundamental enabling technologies, such as vision-based navigation systems which use deep learning to compute obstacle avoidance and hybrid control structures which compromise robot autonomy and centralized control. The synthesis finds a prevailing trend of the integration of the state-of-the-art sensors, Light Detection and Range (LiDAR), Global Positioning System (GPS) and hyperspectral imaging with AI-based decisions to automatize the precision applications in targeted irrigation, harvesting, and early disease identification. These technologies will ensure that there is optimization in the utilization of resources, increased yields, as well as reduced environmental impact. Nevertheless, the analysis also discloses a considerable amount of barriers to the broad adoption, namely, the high initial costs, technical issues in unstructured settings, and insufficient and robust regulatory frameworks. Moreover, the review does not focus only on technical matters but addresses crucial socio-economic and ethical consequences, i.e., changes in labor market, data privacy, and fair access. It is important that future studies focus on the creation of stronger, low cost, and flexed robots. Such directions as improving the collaboration with the swarm of multi-robots, enhancing AI models with simulation-to-reality (sim2real) systems, and setting up policy-aware safety and data governance standards can be identified. The review reiterates the use of intelligent robotics as an instrumental tool in the formation of a productive and sustainable future in global agriculture.
Singh et al. (Sat,) studied this question.