Adamo et al. (2026) Empirical decision model learning for multi-sector greenhouse irrigation under water supply restrictions

Identification

• Journal: Smart Agricultural Technology
• Year: 2026
• Date: 2026-09-09
• Authors: Tommaso Adamo, Lucio Colizzi, Giovanni Dimauro, Emanuela Guerriero, Nunzia Lomonte
• DOI: 10.1016/j.atech.2026.102558

Research Groups

• Department of Engineering for Innovation, University of Salento, Lecce, Italy
• Department of Computer Science, University Aldo Moro, Bari, Italy

Short Summary

This paper proposes an instantiation of the Empirical Decision Model Learning (EDML) paradigm to address the joint Water Volume Allocation (WVA) problem in a multisector greenhouse irrigation system. The framework combines machine learning models with Mixed-Integer Linear Programming (MILP) to optimize WVA decisions across all sectors simultaneously.

Objective

• Develop an empirical decision model that approximates the soil moisture response of each sector to different WVA decisions.
• Integrate the trained machine learning model into a MILP formulation to jointly optimize WVA decisions across multiple sectors under shared hydraulic constraints.

Study Configuration

• Spatial Scale: Multisector greenhouse irrigation system with heterogeneous sectors and distinct water requirements, hydraulic thresholds, and sensitivity to moisture extremes.
• Temporal Scale: Irrigation periods of fixed duration, with the manager committing to a water volume allocation at the beginning of each period.

Methodology and Data

• Models used: Empirical Decision Model Learning (EDML) paradigm combining machine learning models with Mixed-Integer Linear Programming (MILP).
• Data sources: Historical sensor data from IoT soil and environmental sensors, used to train the machine learning model.

Main Results

• The proposed EDML framework achieves a 79.1% reduction in water consumption compared to fixed-interval irrigation.
• 74.4% of sector instances are driven into the optimal agronomic moisture range.
• Solver runtimes below 0.3 s for systems with up to 50 sectors.

Contributions

• This paper contributes to the development of a novel framework for multisector irrigation management under restricted water supply, combining machine learning and MILP to optimize WVA decisions.
• The proposed EDML paradigm provides a flexible and data-driven approach to address the combinatorial nature of multisector irrigation management.

Funding

• Not specified in the paper.

Citation

@article{Adamo2026Empirical,
  author = {Adamo, Tommaso and Colizzi, Lucio and Dimauro, Giovanni and Guerriero, Emanuela and Lomonte, Nunzia},
  title = {Empirical decision model learning for multi-sector greenhouse irrigation under water supply restrictions},
  journal = {Smart Agricultural Technology},
  year = {2026},
  doi = {10.1016/j.atech.2026.102558},
  url = {https://doi.org/10.1016/j.atech.2026.102558}
}