Time-series dataset of non-invasive IoT measurements in Apis mellifera colonies: sunflower pollination service during summer season (2024; Ukraine).

This dataset contains time-series of non-invasive IoT measurements from honey bee (Apis mellifera) colonies monitored in Ukraine during the summer season (01-05-2024 to 31-08-2024). The monitoring period includes a sunflower pollination service window (07-07-2024 to 23-07-2024). Data were collected using AmoHive solar-powered smart hives (project page: https://amohive.com/). Measurements include internal and external temperature (°C), internal and external relative humidity (%), hive weight (kg), and technical parameters: device processor temperature (°C) and voltage after the solar-panel stabilizer (V). The record covers nine monitored hives. RAW telemetry is provided at an approximately hourly acquisition cadence with minor jitter (as exported), while CLEAN data are provided as hourly time series aligned to a fixed local-time grid. Record contents: RAW telemetry files (per hive) as originally exported from the IoT system. CLEAN hourly time series aligned to a fixed local-time grid, including explicit per-cell provenance labels (*_method) for imputed values. A workbook of manually recorded beekeeper events (per hive) and cleaned data files with embedded events. A deterministic processing protocol describing the RAW-to-CLEAN transformation and imputation rules. Study context (sunflower pollination service: 07-07-2024 to 23-07-2024) During sunflower flowering, colonies were organized into two field groups and a stationary apiary (control) group: Field group A: hives placed in a sunflower field. Field group B: hives placed in a sunflower field treated with an ecological attractant. Control group C (stationary apiary): hives kept outside sunflower fields. Hives in field groups A and B were kept at the stationary apiary together with the other colonies before the pollination service and were returned to the stationary apiary after the pollination service. Intended use This dataset supports pollination and colony-behavior research, data quality and imputation studies, and time-series modeling (forecasting, anomaly detection) based on weight, temperature, and humidity signals.