Esteem Professors,
We would like to share our project developed using MATLAB, Raspberry Pi, and an SDS011 sensor as part of our learning in statistical modelling, IoT, and real-world data analysis.
GitHub Repository: (https://github.com/Simran-2400/air-quality-peace-war.git)
Project Overview
The objective was to compare air quality between a peaceful city (Parma, Italy) and a city under active war (Kyiv, Ukraine) during the October 2023 missile attack. We collected 3,600 PM2.5/PM10 readings using a Raspberry Pi + SDS011 sensor in Parma, and used open-access SaveEcoBot ground sensor data for Kyiv. Statistical analysis was performed entirely in MATLAB.
Implemented Features
Time series visualisation of 3,600 readings per city
K-Means clustering (K=3) to classify Clean Air, Moderate, and Extreme pollution events
Elbow Method for optimal K selection
Linear regression PM10 ~ PM2.5 with 95% confidence band
Centroid verification using grpstats() and assert()
Technologies Used
MATLAB (Statistics & Machine Learning Toolbox)
Raspberry Pi + SDS011 sensor
SaveEcoBot open sensor network
GitHub for version control
Key Result
K-Means clustering isolated the October 27th missile attack as an independent extreme cluster — with no labels, no dates, no context provided. PM2.5 peak: 818 µg/m³ — 54× the WHO daily limit.
Teammates:
@AbeerAbuNemer · @ahmedelmenyawe96-ui · @emelian-chkaira
We sincerely thank @MarcoRianiUNIPR , @Asadunipr , and @AldoCorbelliniUNIPR for providing us the Raspberry Pi, their guidance, and for encouraging us to apply our learning to a real-world problem.
Thank you.
Esteem Professors,
We would like to share our project developed using MATLAB, Raspberry Pi, and an SDS011 sensor as part of our learning in statistical modelling, IoT, and real-world data analysis.
GitHub Repository: (https://github.com/Simran-2400/air-quality-peace-war.git)
Project Overview
The objective was to compare air quality between a peaceful city (Parma, Italy) and a city under active war (Kyiv, Ukraine) during the October 2023 missile attack. We collected 3,600 PM2.5/PM10 readings using a Raspberry Pi + SDS011 sensor in Parma, and used open-access SaveEcoBot ground sensor data for Kyiv. Statistical analysis was performed entirely in MATLAB.
Implemented Features
Time series visualisation of 3,600 readings per city
K-Means clustering (K=3) to classify Clean Air, Moderate, and Extreme pollution events
Elbow Method for optimal K selection
Linear regression PM10 ~ PM2.5 with 95% confidence band
Centroid verification using grpstats() and assert()
Technologies Used
MATLAB (Statistics & Machine Learning Toolbox)
Raspberry Pi + SDS011 sensor
SaveEcoBot open sensor network
GitHub for version control
Key Result
K-Means clustering isolated the October 27th missile attack as an independent extreme cluster — with no labels, no dates, no context provided. PM2.5 peak: 818 µg/m³ — 54× the WHO daily limit.
Teammates:
@AbeerAbuNemer · @ahmedelmenyawe96-ui · @emelian-chkaira
We sincerely thank @MarcoRianiUNIPR , @Asadunipr , and @AldoCorbelliniUNIPR for providing us the Raspberry Pi, their guidance, and for encouraging us to apply our learning to a real-world problem.
Thank you.