WaggleNet: A LoRa and MQTT-Based Monitoring System for Internal and External Beehive Conditions

Most smart-hive systems track only internal temperature and humidity and often depend on costly GSM/LTE links. That raises the cost per node for beekeepers monitoring several hives.

This work was conducted during the government-funded IITP Korean AI Square Summer 2025 Program, hosted by Korean Software Square at Purdue University, while I remained affiliated with Jeonbuk National University. The research received guidance from Prof. Eric T. Matson and Prof. Anthony H. Smith, as acknowledged in the paper. WaggleNet uses a three-tier architecture: LoRa worker nodes send data to a LoRa–MQTT gateway, which forwards it to a cloud database and Flutter app. The system records conditions inside the hive alongside the surrounding environment.

The field trial used 2 Langstroth hives fitted with 4 worker nodes (2 internal, 2 external) over two months at the host program's Korean Software Square facilities and a nearby bee yard in West Lafayette.

I contributed to field-test execution, data collection, and evaluation during the two-month trial. I also worked with the international research team to document the methods and results, then co-authored the six-author arXiv preprint as third author.

• Dual-scope sensing: paired internal and external nodes captured hive conditions with surrounding environmental context.
• LoRa at the edge: low-power radio kept worker nodes independent of per-node cellular connections.
• Protocol bridge: the gateway added UTC timestamps and translated LoRa packets into structured MQTT messages.
• Operational feedback: the Flutter client visualized live readings and Firebase Cloud Messaging delivered threshold-based alerts.

Each worker node sampled temperature, humidity, light, and location every three minutes, then transmitted the readings over LoRa. The gateway appended a UTC timestamp and forwarded structured JSON via MQTT to SQLite. The Flutter application displayed readings and geolocation in real time.

Four worker nodes were installed across two Langstroth hives for a two-month field trial. The team measured packet delivery, transmission range, end-to-end latency, uptime, and per-node hardware cost in urban and rural line-of-sight settings.

• Packet delivery: 100% at 110 m urban and 150 m rural line of sight
• Latency: under 5 seconds end to end; 3.8 seconds mean
• Uptime: over 95% across the two-month field trial
• Hardware cost: $25–30 per node, 70–85% below the $200–500 commercial systems used for comparison

The evaluation scope was two hives and four nodes over two months. Separating low-power field transport from cloud messaging kept the field hardware inexpensive while preserving a conventional application interface.