The concept of precision agriculture, which proposes farming management adapted to crop variability, has emerged in recent years. To effectively implement precision agriculture, data must be gathered from the field in an automated manner at minimal cost. In this study, a small autonomous field inspection vehicle was developed to minimise the impact of the scouting on the crop and soil compaction. The proposed approach integrates a camera with a GPS receiver to obtain a set of basic behaviours required of an autonomous mobile robot to inspect a crop field with full coverage. A path planner considered the field contour and the crop type to determine the best inspection route. An image-processing method capable of extracting the central crop row under uncontrolled lighting conditions in real time from images acquired with a reflex camera positioned on the front of the robot was developed. Two fuzzy controllers were also designed and developed to achieve vision-guided navigation. A method for detecting the end of a crop row using camera-acquired images was developed. In addition, manoeuvres necessary for the robot to change rows were established. These manoeuvres enabled the robot to autonomously cover the entire crop by following a previously established plan and without stepping on the crop row, which is an essential behaviour for covering crops such as maize without damaging them.

This paper describes a supervisor system for monitoring the operation of automated agricultural vehicles. The system analyses all of the information provided by the sensors and subsystems on the vehicles in real time and notifies the user when a failure or potentially dangerous situation is detected. In some situations, it is even able to execute a neutralising protocol to remedy the failure. The system is based on a distributed and multi-level architecture that divides the supervision into different subsystems, allowing for better management of the detection and repair of failures. The proposed supervision system was developed to perform well in several scenarios, such as spraying canopy treatments against insects and diseases and selective weed treatments, by either spraying herbicide or burning pests with a mechanical-thermal actuator. Results are presented for selective weed treatment by the spraying of herbicide. The system successfully supervised the task; it detected failures such as service disruptions, incorrect working speeds, incorrect implement states, and potential collisions. Moreover, the system was able to prevent collisions between vehicles by taking action to avoid intersecting trajectories. The results show that the proposed system is a highly useful tool for managing fleets of autonomous vehicles. In particular, it can be used to manage agricultural vehicles during treatment operations.

Many environmental incidents affect large areas, often in rough terrain constrained by natural obstacles, which makes intervention difficult. New technologies, such as unmanned aerial vehicles, may help address this issue due to their suitability to reach and easily cover large areas. Thus, unmanned aerial vehicles may be used to inspect the terrain and make a first assessment of the affected areas; however, nowadays they do not have the capability to act. On the other hand, ground vehicles rely on enough power to perform the intervention but exhibit more mobility constraints. This paper proposes a multi-robot sense-act system, composed of aerial and ground vehicles. This combination allows performing autonomous tasks in large outdoor areas by integrating both types of platforms in a fully automated manner. Aerial units are used to easily obtain relevant data from the environment and ground units use this information to carry out interventions more efficiently. This paper describes the platforms and sensors required by this multi-robot sense-act system as well as proposes a software system to automatically handle the workflow for any generic environmental task. The proposed system has proved to be suitable to reduce the amount of herbicide applied in agricultural treatments. Although herbicides are very polluting, they are massively deployed on complete agricultural fields to remove weeds. Nevertheless, the amount of herbicide required for treatment is radically reduced when it is accurately applied on patches by the proposed multi-robot system. Thus, the aerial units were employed to scout the crop and build an accurate weed distribution map which was subsequently used to plan the task of the ground units. The whole workflow was executed in a fully autonomous way, without human intervention except when required by Spanish law due to safety reasons.

Determining the best path planning for an agricultural task is a very important issue in crop management because it directly affects the distances travelled by the agricultural machines and, accordingly, directly affects the soil compaction that occurs and the inputs that are consumed (time and fuel). However, determining the optimal path is a difficult problem because of the large number of variables that must be taken into account: number of vehicles, speeds, turning radii, geometry of the field, tank capacities, fuel consumption, etc. The problem becomes even more difficult when the vehicles are heterogeneous, that is, they perform at different speeds or have different rates of fuel consumption. Furthermore, the optimal multi-path plan may vary depending on the criteria selected, that is, the paths that are optimal for reducing distance may be very inefficient for saving time. In this paper, a practical approach to multi-path planning for fleets of heterogeneous and autonomous vehicles is proposed, considering the optimisation of several criteria such as distance, time, input cost or some combination thereof.