This systematic review critically evaluates the current state and future potential of real-time, end-to-end smart, and automated irrigation management systems, focusing on integrating the Internet of Things (IoTs) and machine learning technologies for enhanced agricultural water use efficiency and crop productivity. In this review, the automation of each component is examined in the irrigation management pipeline from data collection to application while analyzing its effectiveness, efficiency, and integration with various precision agriculture technologies. It also investigates the role of the interoperability, standardization, and cybersecurity of IoT-based automated solutions for irrigation applications. Furthermore, in this review, the existing gaps are identified and solutions are proposed for seamless integration across multiple sensor suites for automated systems, aiming to achieve fully autonomous and scalable irrigation management. The findings highlight the transformative potential of automated irrigation systems to address global food challenges by optimizing water use and maximizing crop yields.

Accurate estimation of reference evapotranspiration (ET_r) is important for irrigation planning, water resource management, and preserving agricultural and forest habitats. The widely used Penman–Monteith equation (ASCE-PM) estimates ET_r across various timescales using ground weather station data. However, discrepancies persist between estimated ET_r and measured ET_r obtained from weighing lysimeters (ET_r-lys), particularly in advective environments. This study assessed different machine learning (ML) models in comparison to ASCE-PM for ET_r estimation in highly advective conditions. Various variable combinations, representing both radiation and aerodynamic components, were organized for evaluation. Eleven datasets (DT) were created for the daily timescale, while seven were established for hourly and quarter-hourly timescales. ML models were optimized by a genetic algorithm (GA) and included support vector regression (GA-SVR), random forest (GA-RF), artificial neural networks (GA-ANN), and extreme learning machines (GA-ELM). Meteorological data and direct measurements of well-watered alfalfa grown under reference ET conditions obtained from weighing lysimeters and a nearby weather station in Bushland, Texas (1996–1998), were used for training and testing. Model performance was assessed using metrics such as root mean square error (RMSE), mean absolute error (MAE), mean bias error (MBE), and coefficient of determination (R²). ASCE-PM consistently underestimated alfalfa ET across all timescales (above 7.5 mm/day, 0.6 mm/h, and 0.2 mm/h daily, hourly, and quarter-hourly, respectively). On hourly and quarter-hourly timescales, datasets predominantly composed of radiation components or a blend of radiation and aerodynamic components demonstrated superior performance. Conversely, datasets primarily composed of aerodynamic components exhibited enhanced performance on a daily timescale. Overall, GA-ELM outperformed the other models and was thus recommended for ET_r estimation at all timescales. The findings emphasize the significance of ML models in accurately estimating ET_r across varying temporal resolutions, crucial for effective water management, water resources, and agricultural planning.

Site-specific measurements of the crop yield during harvesting are essential for successfully implementing precision management techniques. This study aimed to estimate the mass of radish tubers using the impact principle under simulated vibration and sloped-field harvesting conditions with a laboratory test bench. These conditions included the conveyor speed (CS), impact plate layout (IP), falling height onto the impact plate (FH), the plate angle relative to the horizontal (PH), the field slope, and the vibration of the harvesting machine. Two layouts of impact-type sensors were fabricated and tested, one with a single load cell (SL) and the other with two load cells (DL). An adjustable slope platform and a vibration table equipped with vibration blades were utilized to simulate the slope and vibration effects, respectively. Calibrations were conducted to verify the accuracy of the sensor outputs, processed with the finite impulse response and moving average filters. Radish mass was estimated using an asymmetrically trimmed mean method. The relative percentage error (RE), standard error (SE), coefficient of determination (R²), and analysis of variance (ANOVA) were used to assess the impact plate performance. The results indicated that the SE for both impact plates was less than 4 g in the absence of vibration and slope conditions. The R² for the single and double impact plates ranged from 0.58 to 0.89 and 0.69 to 0.81, respectively. The FH had no significant impact, while the PH significantly affected the mass measurements for both impact plates. On the other hand, the CS significantly affected the plate performance, except for the double-load-cell impact plate. Both vibration and slope affected the mass measurements, with RE values of 9.89% and 13.92%, respectively. The RE for filtered radish signals was reduced from 9.13% to 5.42%. The tests demonstrated the feasibility of utilizing the impact principle to assess the mass of radishes, opening up possibilities for the development of yield-monitoring systems for crops harvested in a similar manner.