Hasan et al. (2025) Bangladesh IoT Environmental Dataset (BIED)

Identification

• Journal: Mendeley Data
• Year: 2025
• Date: 2025-11-06
• Authors: Hasan, Md Mehedi, Rashid, Maria
• DOI: 10.17632/956gt333hk.1

Research Groups

• Daffodil International University

Short Summary

The Bangladesh IoT Environmental Dataset (BIED) is a curated collection of simulated IoT sensor readings representing diverse environmental conditions across Bangladesh, designed to support AI, IoT, and environmental research through analysis, prediction, and system validation.

Objective

• To create and provide a realistic, context-specific IoT environmental dataset for Bangladesh, simulating continuous monitoring of critical environmental parameters, to facilitate research in AI, IoT, and environmental monitoring, prediction, and system validation.

Study Configuration

• Spatial Scale: Five distinct regions across Bangladesh: Urban (Dhaka), Rural (Rajshahi), Coastal (Khulna), Industrial (Narayanganj), and Agricultural (Bogura).
• Temporal Scale: Hourly readings simulated over several days, incorporating realistic diurnal, seasonal, and regional variations.

Methodology and Data

• Models used: Not applicable; the paper describes a dataset generated through simulation rather than using pre-existing models for analysis. The dataset itself is generated using "Sensor Behavior Modeling" to include realistic drift, noise, and fluctuation patterns.
• Data sources: Simulated IoT sensor readings. The dataset is a synthetic creation mimicking real-world sensor data, not derived from external observational or reanalysis sources.

Main Results

• The Bangladesh IoT Environmental Dataset (BIED) was developed, comprising over 10,000 records in CSV format.
• The dataset includes seven environmental variables: Temperature (°C), Humidity (%), Air Quality Index (AQI), Carbon Dioxide (CO₂) levels (ppm), Soil Moisture (%), Rainfall (mm), and Light Intensity (lux).
• Data is provided with hourly temporal resolution and covers five distinct regional contexts within Bangladesh.
• The dataset incorporates realistic sensor behavior modeling, including drift, noise, and fluctuation patterns, to enhance its applicability for real-world system validation.

Contributions

• Provides the first known Bangladesh-contextualized IoT environmental dataset, addressing a gap in regionally specific data for AI, IoT, and environmental research.
• Offers a versatile resource for various applications, including smart agriculture, smart city systems, climate research, anomaly detection in IoT systems, AI/TinyML model training, and IoT platform testing.
• Enhances realism by incorporating simulated sensor drift, noise, and fluctuations, making it suitable for validating robust IoT systems and algorithms.

Funding

• Not specified in the provided text.

Citation

@article{Hasan2025Bangladesh,
  author = {Hasan, Md Mehedi and Rashid, Maria},
  title = {Bangladesh IoT Environmental Dataset (BIED)},
  journal = {Mendeley Data},
  year = {2025},
  doi = {10.17632/956gt333hk.1},
  url = {https://doi.org/10.17632/956gt333hk.1}
}