Research library

Time-series dataset of honey bee colony dynamics before, during, and after sunflower pollination

Open 2026 BMC Research Notes data paper releasing synchronized smart-hive IoT time series from nine Apis mellifera colonies around a sunflower pollination-service window.

Publication details

Organizations
🇺🇦 AmoHive / Ukraine smart-hive deployment
Year
2026
Type
Dataset

[PDF](/assets/research/papers/pdfs/Time-series dataset of honey bee colony dynamics before during and after sunflower pollination.pdf)

Abstract

Objectives

Precision beekeeping is an integral part of precision agriculture, which relies on sensor technologies and high-quality datasets to quantify and optimize ecosystem services such as crop pollination. To support reproducible research and the planning and evaluation of crop pollination campaigns in precision beekeeping, we release a time-series dataset that characterizes colony dynamics before, during, and after pollination, using sunflower as a case study.

Data description

We release synchronized, non-invasive time series from nine smart hives (Apis mellifera) monitored in Ukraine (Europe/Kyiv) from 01 May to 31 Aug 2024, including a sunflower pollination service window (07–23 Jul 2024) and a documented attractant intervention. Sensors record hive weight, in-hive and ambient temperature, in-hive and ambient relative humidity, and device signals (processor temperature and stabilized solar voltage). The repository includes raw telemetry exports, cleaned hourly series aligned to a fixed local time grid, and a beekeeper event log, together with reproducible scripts and a documented processing protocol.

Relevancy to Gratheon

This dataset paper is valuable for Gratheon's time-series analytics because it provides real smart-hive telemetry with synchronized weight, in-hive/ambient temperature, humidity, power/device status, event logs, and cleaned hourly series. The pollination-service framing is especially relevant for future Gratheon features around colony readiness, pollination-contract monitoring, anomaly detection, sensor-data cleaning, and forecasting. The Zenodo repository and documented preprocessing pipeline make it a practical benchmark for backend data models and dashboard experiments.