Data Driven Worker Routing for Irrigation Monitoring

Abstract

Survival and thrift in a globally competitive market strongly depend on the improvement of production quality while monitoring the pricing for grafting nurseries. Real-time monitoring of irrigation quality is a necessary yet cumbersome task for propagation facilities. Even for indoor facilities with automatic irrigation systems, this task remains bothersome due to the large greenhouses. The main goal of this paper is to propose an irrigation monitoring system to increase production productivity by minimizing the time workers spend on responding to irrigation monitoring calls. Here, a two-stage data-driven approach is proposed, where the traveling salesman model is implied on a reduced spatiotemporal network provided by Bayesian analysis to optimize workers routing for irrigation monitoring. The proposed approach is the final piece of a dynamic data-driven simulation-based optimization framework. The results over synthetic data demonstrate a potential reduction of 11.60% in traveled distance per worker per call.