Published June 12, 2026 | Version v1

K-HEALTHinAIR — Pilot DE (Germany): Indoor Air Quality Analysis

  • 1. ROR icon Atos (Spain)

Description

# K-HEALTHinAIR — Pilot DE (Germany): Indoor Air Quality Analysis

## Overview

This dataset and set of analysis notebooks correspond to the **German pilot** of the K-HEALTHinAIR EU-funded project. The pilot monitors indoor and outdoor air quality in two building types: **university lecture halls** (LEC) and **canteens** (CAN), with campaigns running from March 2023 to December 2024.

These analyses are intended to be used with data collected from sensors as deployed in the pilot sites. The data provided is the subset filtered by pilot owners from the K-HEALTHinAIR platform to carry out specific analysis.

This pilot places particular emphasis on **time-series forecasting** (comparing ARIMA, SARIMA, Random Forest, LSTM, and Temporal Convolutional Networks) and on the application of a **GO-AQS health scoring framework** that classifies air quality against evidence-based health thresholds.

---

## Contents

```
DE/
├── README.md
├── EDA_DE_LEC.ipynb
├── EDA_DE_LEC_Aggregation.ipynb
├── EDA_DE_LEC_Aggregation_Outliers.ipynb
├── EDA_DE_LEC_Outliers.ipynb
├── EDA_DE_LEC_Outliers_Methodology.ipynb
├── EDA_DE_CAN_corr_I_O.ipynb
├── EDA_DE_CAN_Outliers.ipynb
├── EDA_DE_corr_I_O_file.ipynb
├── Forecasting_DE_LEC.ipynb
├── GO_AQS_DE_CAN.ipynb
├── GO_AQS_DE_LEC.ipynb
└── Data/
    ├── DECAN_*.csv    (4 files — canteens)
    └── DELEC_*.csv    (3 files — lecture halls)
```

---

## Notebooks

### `EDA_DE_LEC.ipynb`
Exploratory analysis of lecture hall sensor data: descriptive statistics, boxplots, kernel density estimation, and Pearson correlation matrices across indoor parameters.

### `EDA_DE_LEC_Aggregation.ipynb`
Hourly and daily aggregation of raw lecture hall measurements. Generates the aggregated datasets used in downstream analyses.

### `EDA_DE_LEC_Aggregation_Outliers.ipynb`
Combines aggregation with outlier detection on lecture hall data, applying Isolation Forest to flag anomalous measurements.

### `EDA_DE_LEC_Outliers.ipynb`
Dedicated outlier analysis for lecture halls using Isolation Forest anomaly scoring and decision function visualisation.

### `EDA_DE_LEC_Outliers_Methodology.ipynb`
Methodological exploration of outlier detection approaches for lecture hall data, documenting parameter sensitivity and threshold selection.

### `EDA_DE_CAN_corr_I_O.ipynb`
Indoor–outdoor correlation analysis for canteen sensors: Pearson heatmaps, I/O ratio computation, and multi-area comparison.

### `EDA_DE_CAN_Outliers.ipynb`
Outlier detection for canteen data using Isolation Forest. Includes normality (Shapiro–Wilk) and stationarity (ADF) testing.

### `EDA_DE_corr_I_O_file.ipynb`
File-level indoor–outdoor correlation analysis combining canteen and lecture hall scenarios.

### `Forecasting_DE_LEC.ipynb`
Multi-model time-series forecasting of indoor air quality in lecture halls. Models compared: ARIMA, SARIMA, Autoregressive + Random Forest, LSTM (Long Short-Term Memory), and Temporal Convolutional Network (TCN). Includes seasonal decomposition, stationarity testing (ADF), lag plots, and autocorrelation analysis.

### `GO_AQS_DE_CAN.ipynb`
Application of the GO-AQS (Global Outdoor Air Quality Score) framework to canteen data. Classifies PM₂.₅ and CO₂ levels into health categories: Good (PM₂.₅ ≤ 8 µg/m³; CO₂ ≤ 850 ppm), Moderate, and Unhealthy.

### `GO_AQS_DE_LEC.ipynb`
Same GO-AQS health scoring applied to lecture hall data.

---

## Data Files

### Column structure

All raw CSV files share the following columns:

| Column | Description | Units |
|---|---|---|
| `_time` | Measurement timestamp (hourly resolution) | UTC datetime |
| `sub_scenario` | Sub-scenario code (e.g., `CAN01`, `LEC01`) | — |
| `area` | Area identifier (e.g., `CAN01001`, `LEC01002`) | — |
| `sensor_id` | Anonymised sensor identifier (UUID) | — |
| `co2` | Indoor CO₂ | ppm |
| `humidity` | Indoor relative humidity | % |
| `pm1` | Indoor PM₁ | µg/m³ |
| `pm10` | Indoor PM₁₀ | µg/m³ |
| `pm25` | Indoor PM₂.₅ | µg/m³ |
| `temperature` | Indoor temperature | °C |
| `tvocs` | Indoor total VOCs | ppb |
| `out_co` | Outdoor CO | mg/m³ |
| `out_humidity` | Outdoor relative humidity | % |
| `out_no2` | Outdoor NO₂ | µg/m³ |
| `out_o3` | Outdoor O₃ | µg/m³ |
| `out_pm10` | Outdoor PM₁₀ | µg/m³ |
| `out_pm25` | Outdoor PM₂.₅ | µg/m³ |
| `out_so2` | Outdoor SO₂ | µg/m³ |
| `out_temperature` | Outdoor temperature | °C |

Outdoor values are shared across all sensors in the same area (single reference station per area). Missing values are represented as empty cells.

### File inventory

| File | Period | Scenario |
|---|---|---|
| `DECAN_2023-03-01_2023-11-15.csv` | Mar–Nov 2023 | Canteens |
| `DECAN_2023-09-01_2024-01-16.csv` | Sep 2023 – Jan 2024 | Canteens |
| `DECAN_2023-11-13_2024-02-12.csv` | Nov 2023 – Feb 2024 | Canteens |
| `DECAN_2024-01-01_2024-12-19.csv` | Jan–Dec 2024 | Canteens |
| `DELEC_2023-03-01_2023-11-16.csv` | Mar–Nov 2023 | Lecture halls |
| `DELEC_2023-10-01_2024-01-25.csv` | Oct 2023 – Jan 2024 | Lecture halls |
| `DELEC_2024-01-01_2024-12-19.csv` | Jan–Dec 2024 | Lecture halls |

Multiple overlapping files for the same scenario represent different extraction windows used at different stages of analysis; they are not exact duplicates.

---

## Requirements

Python 3.10+. Main dependencies:

```
pandas, numpy, matplotlib, seaborn, scipy, scikit-learn,
statsmodels, tensorflow, openpyxl
```

---

## License

Copyright (C) 2021–2026 Advanced Computing Spain SL – BULL.
Released under the **GNU General Public License v3.0**.
See <https://www.gnu.org/licenses/> for the full text.

---

## Acknowledgements

This work was carried out within the K-HEALTHinAIR project, funded by the European Union. Views expressed are those of the authors only and do not necessarily reflect those of the European Union or the funding agency.

Files

German pilot.zip

Files (88.0 MB)

Name Size
md5:5493c5dd77409914194fc85ce7dfa1cc
88.0 MB Preview Download

Additional details

Funding

European Commission
K-HEALTHinAIR - Knowledge for improving indoor AIR quality and HEALTH 101057693

Software

Programming language
Python