Conference Agenda

Ion-selective field-effect transistors (ISFETs) have great potential for the real-time monitoring of ions in aqueous solutions.

This article presents application-specific examples of ISFET-based measurement systems that have been implemented in two agricultural applications: in-field measuring of soil nutrients and determining nutrient concentration in hydroponic systems. To this end, commercially available ISFETs have been incorporated into multi-sensor modules (lab-on-chip) or as standalone sensors to detect nutrients available to plants, such as H₂PO₄⁻, NO₃⁻, K⁺, NH₄⁺ and pH. In addition, relevant temperatures were measured for calibration purposes. The ISFETs used were selected according to the application's specific requirements in terms of measurement dynamics. A laboratory test bench was set up for validation purposes. Using suitable reference measurements, the sensors' output signals were analysed in the laboratory with regard to stability against disturbance variables, reproducibility, and transient behaviour. In addition to long-term stability, the conditioning and calibration procedures of the ISFET were identified as important factors. A baseline drift correction in the form of continuous single-point calibration was also developed and successfully implemented. ISFET-based measurement systems were integrated into newly developed, application-specific mechatronic handling systems and process sequences.

This article presents the implementation and experimental validation of these systems, including the first results.

In hydroponics, nutrient solution reuse has been widely adopted to improve nutrient use efficiency and reduce environmental impacts. In closed hydroponic systems, selective ion uptake continuously alters the ionic composition of the nutrient solution, limiting the applicability of EC (Electrical Conductivity)-based management for reuse. To address this limitation, ion-sensing-based nutrient control approaches have been proposed, enabling direct monitoring of individual ion concentrations and nutrient supply through variable adjustment of fertilizer inputs. However, strategies that control all individual nutrients require numerous fertilizer tanks and complex control schemes, increasing system complexity, reducing space efficiency, and raising operational demands. In addition, complete and stable sensing of relevant ions remains challenging under practical operating conditions. Therefore, nutrient management systems require a simplified control framework based on monitoring a minimal number of representative ions while maintaining sufficient control performance. In this study, a nitrate-centered nutrient control strategy is proposed considering the environmental impact of nitrate discharge. The approach improves nutrient solution management compared to EC-based methods while remaining simpler than ion-specific control strategies. The system uses four fertilizer tanks structured around different nitrate contributions, while NO₃⁻, K⁺, and Ca²⁺ concentrations are monitored using SCISEs, and an algorithm determines tank combinations to achieve target ion concentrations.

Climate change is progressively reducing water availability for irrigation in Mediterranean regions, requiring more efficient, adaptive and resilient management strategies. This study presents an innovative framework for the full automation of water distribution in gravity-fed irrigation networks, aimed at improving water-use efficiency at both field and district scales. The system was tested in a gravity-fed irrigation district in the Mantua area (northern Italy), within the basin managed by the Garda Chiese irrigation agency.

The framework introduces precision surface irrigation technologies into a traditionally surface-irrigated district. It includes interconnected automated sluice gates installed at key control nodes of the network and at field inlets, managed through dedicated software that coordinates gate operations in real time. Gates and software were developed by the University of Milan in collaboration with the Garda Chiese irrigation agency and Digital Drop, a university spinoff, demonstrating effective cooperation among academia, local stakeholders and private SMEs. A digital booking platform collects farmers’ irrigation requests, replacing fixed rotation schedules with flexible, demand-driven allocation.

Performance was assessed over two irrigation seasons by analyzing water volumes, hydrodynamics, conveyance losses, crop yield and quality. Results showed water savings of up to 30%, confirming the potential of fully automated, demand-driven management in surface-irrigated districts.

Climate change, as well as water scarcity, is a major constraint to the sustainable production of watermelon by smallholder farmers in the context of developing countries. This study presents the design, development, and validation of a cost-effective precision drip irrigation model that is informed by the climate smart agriculture paradigm. The precision drip irrigation model was designed based on crop evapotranspiration, soil physical properties, as well as the optimal spacing of emitters to ensure the maximum uniformity of water distribution. The performance of the precision drip irrigation model was compared with the conventional surface irrigation model using a comparative experimental study. The performance indicators of the precision drip irrigation model were evaluated using the emission uniformity, distribution uniformity, water use efficiency and yield response. The findings of the study show that the precision drip irrigation model that was developed can attain an emission uniformity of over 85%, reduce the rate of water application by 35-40%, as well as increase the fruit yield by over 20% compared to the conventional irrigation model. Also, the water productivity of the precision drip irrigation model shows a significant increase. This model can be used as a tool towards the sustainable watermelon production.

Adaptive crop regulation requires integrating plant physiological processes with precise environmental control.
This study presents a plant‑based irrigation and nutrient management concept that links photosynthetic activity, water uptake, and nutrient fixation through high‑resolution measurement of irrigation supply, drain volumes and water quality. An adaptive system approach, enables construction of a dynamic mass balance that reflects real-time plant responses to light, climate, and crop-handling conditions.

The framework recognizes the dominant role of water in plant functioning. By applying adjustable moving averages, the system accounts for circadian rhythms and developmental changes in leaves, flowers, and fruits. This adaptive control strategy minimizes root stress, stabilizes physiological processes, and improves the predictability of yield and crop quality. The approach is designed for integration into existing control systems, offering a pathway toward more responsive, physiology‑driven greenhouse and (open field) irrigation management.

The global hydroponics market is expanding rapidly. However, microplastics (MPs) have emerged as a new threat to crop safety, as their uptake through plant roots has been reported to induce phytotoxic responses. Despite this concern, studies directly detecting and quantifying MPs in nutrient solutions remain scarce. In this study, visible-near-infrared and short-wave infrared hyperspectral imaging (VNIR: 400–1000 nm; SWIR: 900–1700 nm) was employed to identify MP types and quantify concentrations across four liquid matrices, including distilled water and three agricultural nutrient solutions. Machine learning and deep learning-based multi-label classification was performed on three individual polymers and their mixtures, achieving an exact match ratio exceeding 0.90. The limit of detection was determined through concentration-gradient experiments, and cross-matrix transfer validation confirmed that models could classify MPs under unseen nutrient solution conditions. Explainable AI (XAI) analysis verified that key discriminative wavelengths were consistent with known polymer absorption characteristics. This study demonstrates that hyperspectral imaging combined with machine learning and deep learning is an effective non-destructive tool for MP identification and quantification in nutrient solutions, contributing to systematic MP monitoring in hydroponic systems.

Acknowledgement: This research was supported by a grant (23192MFDS106) from Ministry of Food and Drug Safety in 2026.