Field robots for precision agriculture

Abstract

In this review, a widely implemented method was followed. Therefore, the literature review process was divided into three phases: Planning, execution, and result analysis. Precision agriculture is a management strategy that takes into account the temporal and spatial variability to improve sustainability of agricultural production. The precision agriculture cycle is constituted by three stages: Geo-referenced measurement of within-field parameters; analysis and interpretation of geo-referenced data for mapping within-field parameters; and spatially variable rate crop input application. The instruments and techniques needed for implementing precision agriculture are Global Navigation Satellite Systems (GNSS), proximal and/or remote sensing, spatially-based software, soil-crop simulation models, controllers for spatially variable rate crop input application, guidance systems of agricultural machines, and field robots. During the precision agriculture cycle, instead of unmanned tractors or together with them, it is possible to use field robots for seeding and planting, plant protection, fruit harvest, and other crop operations. As a consequence of intensification, mechanisation, and automation, agricultural production has significantly increased over time. In both traditional and precision agriculture, the automation of crop operations is possible through the employment of robots, as they can accomplish repetitive labour tasks by keeping high precision, as well as saving time and energy during the working day. A total of 20 case studies of implementation of robots in agriculture and forestry were counted: Optimised coverage for arable farming; weed control; high precision seeding; crop yield estimation; precision irrigation; tree fruit production; vehicle formation control; date palm tree spraying; plant probing; cucumber harvesting; cucumber leaf removal; rose harvesting; strawberry harvesting; pot handling in nurseries and greenhouses; precision forestry; semi-automation of forwarder crane; livestock breeding and nurturing; livestock exploitation; livestock harvesting, slaughtering and processing; and aerial-based precision agriculture. In this review, Unmanned Ground Vehicles (UGVs) were classified according to different parameters and some examples were described for each category of agricultural UGVs or field robots. Precision agriculture will be widely implemented if cost-effective field robots are developed. From this study, it is possible to conclude that the most researched perception solutions are based on vision and cloud point sensors, and the UAV carrying some sensors is the preferred robotic solution for monitoring a large field, while a ground-based robot shows a unique design based on its required tasks. As such, agricultural tasks are becoming increasingly automated, above all in high-profit agriculture.