IoT-Based Automation of a Reverse-Osmosis Desalination Process in the Galápagos Islands

Reliable drinking-water production is difficult on remote islands where brackish-water delivery is intermittent, technical personnel are scarce, and reverse-osmosis plants are manually operated. This study presents an operational characterization of an Internet of Things (IoT) retrofit deployed in Santa Cruz, Galápagos; it is not a controlled before-and-after effectiveness evaluation. An ESP32-based M5Stack Tough controller, distributed ESP-NOW sensing nodes, relay–contactor interfaces, a binary pressure permissive, and a ThingSpeak cloud layer were integrated without replacing the existing pumps and membranes. The exported primary-flow channel contained 4,603,989 numeric observations, including 500 pre-official test readings. Operational analyses used 4,603,489 numeric observations from the official monitoring period; 4,603,340 values remained after nominal-range filtering, and positive flow had a median of 12 L/min (interquartile range: 11–15 L/min). Among 332 logged high-pressure commands, 326 were preceded by a low-pressure command (98.2% unbounded command-state consistency), whereas 275 occurred within a 120 s analytical bound (82.8%). The median low-to-high command delay was 27 s (interquartile range: 11–70 s). Four organizational representatives completed a published 41-item Industry 4.0 maturity instrument before and after deployment; the self-reported overall mean was 0.26 at baseline and 1.95 post-deployment, and these results are interpreted descriptively. Energy-consumption and production data were confidential and unavailable to the authors, while water-quality variables were not measured. The contribution is therefore a long-duration, local-first legacy retrofit with auditable telemetry and explicit limitations, rather than a claim of optimized desalination performance.