Design and Field Evaluation of a Low-Cost, Solar-Powered IoT Node for Digital McPhail Traps in Olive Groves

Monitoring the olive fruit fly (Bactrocera oleae) over large areas needs inexpensive, low-power traps that run unattended for months. We present a low-cost (∼EUR 250), solar-powered IoT node that retrofits a McPhail trap into a Digital McPhail Trap. A Raspberry Pi 3B+ logs temperature and humidity every 15 min, captures three daily images and powers a 4G modem ∼1.5 h/day, and bench measurements give an indicative ≈0.93 W (a probable lower bound). Across six units on Lefkada, Greece, in 2022 (526 device-days), two ran ∼3 months at ∼98% uptime while two stopped until sunset on almost half the days (fleet mean 78.5%). The node tracked temperature trends but is not a reference-grade microclimate instrument. Daily means followed the ERA5-Land consistency reference (r=0.91–0.97), but hourly readings showed a common diurnal discrepancy (RMSE 3.6 °C), reduced by 39% with a leave-one-trap-out correction, and humidity was unusable. A proof-of-concept counter reached r=0.84 within a trap but 0.53 on an unseen one. Catches showed a suggestive negative temperature trend, sensitive to season and autocorrelation and not an established effect. On-trap temperature gave no advantage over reanalysis, and degree-days did not explain between-site differences. This six-site, single-season study demonstrates engineering feasibility, not performance across regions, seasons or trap types.