Team Vayu, consisting of Nitin Nilesh (MS, CSE); Sara Spandhana (MS, ECE); Om Kathalkar and Shreyas Gujar (Interns, College Research Affiliate Program – IoT) from SPCRC and SCRC won the Environment Sensing Project Competition (2022) while team AirIoT comprising Ayu Parmar (MS, ECE); Nitin Nilesh (MS, CSE) and Ritik Yelekar (Intern, College Research Affiliate Program – IoT) from SPCRC and SCRC, made it to the short-list. The event was organized by the MegaSense team, University of Helsinki. The project competition was open to students, and participants are expected to design, build, deploy, and demonstrate an end-to-end solution for environmental sensing. Out of the several teams from India and Finland, eight were short-listed. The competition was in December 2022 but the list was made public in February.
Team Vayu explains their project, IoT-based AQI Estimation using Image Processing and Learning Methods here: This experiment was conducted to provide an IoT-based real-time air quality index (AQI) estimation technique using images and weather sensors in the Indian city of Hyderabad. A mixture of image features, i.e., traffic density, visibility, and sensor features, i.e., temperature and humidity, were used to predict the AQI. Object detection, localization-based Deep Learning (DL) methods, and image processing techniques were used to extract image features. At the same time, a Machine Learning (ML) model was trained on those features to estimate the AQI. In order to conduct this experiment, a dataset containing 5048 images and co-located AQI values across different seasons was collected by driving on the roads of Hyderabad in India. The experimental results report an overall accuracy of 82% for AQI prediction.
Full paper: https://openreview.net/forum?id=LF5JtTD3R7
Youtube link: https://www.youtube.com/watch?v=O0D9UNXolio
Team AirIoT implemented a new method to measure the air quality index (AQI) in real-time using IoT and machine learning (ML) based on traffic data. Team AiIoT explains their project, IoT and ML-based AQI estimation using Real-time Traffic Data here: By deploying PM monitoring nodes in 15 different locations with diverse traffic conditions on Indian roads and collecting a large traffic dataset, the study trains three ML models (random forest, support vector machine, and multi-layer perceptron) to predict AQI into five categories. Additionally, the ML models were trained on individual node datasets to observe the behaviour of AQI levels.
Full paper: https://openreview.net/forum?id=p-btxSKN8v
Youtube link: https://www.youtube.com/watch?v=bAP4GQgruRE
More details about Environment Sensing Project Competition: