Pilot Case 4
Robotic technologies for crop monitoring and management in soilless tomato cultivation in Italy
Location: Experimental and commercial settings in Italy
Overview
Pilot Case 4 focuses on crop monitoring and management in soilless tomato cultivation in Italy, with current activities centered on the integration and validation of sensing and computing components for greenhouse monitoring. During the reporting period, the sensor suite, battery system, and onboard computing units were integrated into a mock-up cart that serves as a simulated mobile support platform for testing sensing, monitoring, and data acquisition capabilities. The system is intended to support both continuous crop observation and channel specific fertigation management. A first greenhouse data acquisition campaign was completed during the tomato growing cycle from October 2025 to January 2026 in Mola di Bari, Italy, and a second campaign started in April 2026. Based on the collected dataset, a first deep learning based processing pipeline has been developed to extract plant structural features, such as plant height, and fruit maturity information.
Timeline
The pilot has entered an implementation and early validation phase, with sensing and computing components integrated into a representative greenhouse monitoring setup. Initial activities included system integration and a first full-cycle experimental campaign, followed by the launch of a second campaign in April 2026. The next phase will focus on extended greenhouse validation, optimization of segmentation and phenotyping algorithms, integration with existing cultivation infrastructures such as rail-mounted platforms or guided carts, dissemination of first results, and publication of selected datasets through a dedicated GitHub repository.
Validation
Validation is being carried out through experimental greenhouse campaigns and comparison with conventional agronomic practices. Results from the first campaign were used to train the developed algorithms and to assess the feasibility of vision-based monitoring by comparing outcomes with traditional manual data gathering. Upcoming validation activities will focus on continuous monitoring performance, robustness of data acquisition, infrastructure integration, and the use of monitoring outputs to support channel-specific fertigation management.
AR Integration
Operators will use AR interfaces to view crop health in immersive formats and engage with interactive training content to enhance their skills in managing soilless tomato systems using robotics.