Anonymised Indoor Air Quality Benchmarking Dataset -- Manufacturing Facilities

Description

This dataset contains anonymised benchmarking indicators derived from indoor air‑quality sensor measurements collected from manufacturing facilities participating in the Smart Manufacturing Data Hub (SMDH) programme.

The dataset summarises air‑quality performance across multiple companies and sectors using aggregated environmental indicators and statistical benchmarking metrics. The purpose of the dataset is to enable cross‑company environmental performance comparison while preserving company anonymity.

The dataset focuses on indoor air‑quality conditions derived from CO₂ and Total Volatile Organic Compounds (TVOC) sensor measurements. These measurements are converted into air‑quality indices and benchmarking indicators that allow comparison across facilities with different operational characteristics.

Data Source

The data originates from IFM environmental sensors installed in participating manufacturing facilities. Raw sensor measurements were processed and aggregated before inclusion in this anonymised dataset.

Anonymisation

To protect company confidentiality:

• Company names are replaced with anonymised identifiers (company1, company2, etc.).
• Sector labels are represented as coded identifiers (sector_code).
• No facility‑level identifiers or timestamps that could reveal operational patterns are included.
• The dataset contains aggregated benchmarking metrics rather than raw sensor readings.

Dataset Structure

Each row represents a company‑level observation for a specific time period, including environmental indicators and benchmarking metrics used to compare air‑quality performance across companies.

The dataset includes:

• Environmental performance indicators
• Normalised benchmarking metrics
• Data coverage indicators
• Statistical benchmarking metrics (z‑scores)
• Relative performance classifications

Analytical Methodology

1. Sensor Data Preparation
   • Cleaning and validation of IFM sensor measurements
   • Removal of missing or inconsistent observations
2. Indicator Construction
   • Conversion of CO₂ and TVOC measurements into air‑quality index (AQI) metrics
3. Benchmarking Normalisation
   • Normalisation of AQI metrics to enable cross‑company comparison
4. Coverage Assessment
   • Calculation of data availability metrics to ensure fair benchmarking
5. Statistical Benchmarking
   • Calculation of z‑scores to quantify relative performance compared with peer companies
6. Performance Classification
   • Classification of company performance as Better, Average, or Worse relative to peers

Intended Use

This dataset supports:

• Cross‑company environmental benchmarking
• Manufacturing environmental performance analysis
• Indoor air‑quality research
• Development of environmental performance dashboards
• Academic or industrial research into workplace environmental conditions

Limitations

• The dataset contains derived benchmarking indicators rather than raw sensor readings.
• Results should be interpreted as relative benchmarking indicators rather than regulatory compliance assessments.
• Sector codes are anonymised and do not represent specific industries.