Modern agriculture is related to a revolution that occurred in a large group of technologies (e.g., informatics, sensors, navigation) within the last decades. In crop production systems, there are field operations that are quite labour-intensive either due to their complexity or because of the fact that they are connected to sensitive plants/edible product interaction, or because of the repetitiveness they require throughout a crop production cycle. These are the key factors for the development of agricultural robots. In this paper, a systematic review of the literature has been conducted on research and commercial agricultural robotics used in crop field operations. This study underlined that the most explored robotic systems were related to harvesting and weeding, while the less studied were the disease detection and seeding robots. The optimization and further development of agricultural robotics are vital, and should be evolved by producing faster processing algorithms, better communication between the robotic platforms and the implements, and advanced sensing systems.

The pulse width modulation (PWM) spray system is the most advanced technology to obtain variable rate spray application without varying the operative sprayer parameters (e.g. spray pressure, nozzle size). According to the precision agriculture principles, PWM is the prime technology that allows to spray the required amount where needed without varying the droplet size spectra which benefits both the uniformity of spray quality and the spray drift reduction. However, some concerns related to the effect of on–off solenoid valves and the alternating on/off action of adjacent nozzles on final uneven spray coverage (SC) have arisen. Further evaluations of PWM systems used for spraying 3D crops under field conditions are welcomed. A tower-shaped airblast sprayer equipped with a PWM was tested in a vineyard. Twelve configurations, combining duty cycles (DC: 30, 50, 70, 100%) and forward speeds (FS: 4, 6, 8 km h⁻¹), were tested. Two methodologies, namely field-standardized and real field conditions, were adopted to evaluate the effect of DC and FS on (1) SC variability (CV%) along both the sprayer travel direction and the vertical spray profile using long water sensitive papers (WSP), and (2) SC uniformity (I_U, index value) within the canopy at different depths and heights, respectively. Furthermore, the SC (%) and deposit density (Nst, no stains cm⁻²), determined using short WSP, were used to evaluate the spray application performances taking into account the spray volumes applied. Under field-controlled conditions, the pulsing of the PWM system affects both the SC variability measured along the sprayer travel direction and along the vertical spray profile. In contrast, under real field conditions, the PWM system does not affect the uniformity of SC measured within the canopy. The relationship between SC and N_st allowed identification of the ranges of 200–250 and 300–370 l ha⁻¹ as the most suitable spray volumes to be applied for insecticide and fungicide plant protection products, respectively.

The H2020-project OPTIMA concept of smart sprayer relies on several functionalities, including variable nozzle spacing for bed-grown carrots, based on an air-assisted boom sprayer. A prototype boom was designed and evaluated though canopy deposition, ground losses, and spray drift potential. Four bed spray configurations, including various nozzle types, angles, and sizes (XR8004, combination of AIUB8504/AI11004, AI8004, and XR8002) at the most appropriate nozzle spacing and height, were tested and compared to a broadcast application (XR11004). Deposition measurements were performed on carrots in bins at early and full-grown stages with respective target zone width of 1.4 m and 2.2 m. Spray drift potential measurements were performed following ISO 22401, 2015. The spray boom was equipped with an air sleeve providing different air speeds (0, 4, 8 m s⁻¹). The relative depositions at both growth stages showed a significant effect of spray configuration and lowest values were found for the broadcast application. The configurations consisting of air inclusion nozzles generated the lowest drift potential compared to the broadcast application, although not significantly different. Bed spray configurations can thus improve canopy depositions and spray drift potential compared to a conventional broadcast application when the boom height and the nozzle spacing are adjusted to the growth stage.