Particulate Matter Dataset Collected with Vehicle Mounted IoT Devices in Delhi-NCRDownload PDF

08 Jun 2021 (modified: 24 May 2023)Submitted to NeurIPS 2021 Datasets and Benchmarks Track (Round 1)Readers: Everyone
Keywords: air pollution, particulate matter, mobile monitoring, interpolation, anomaly detection
TL;DR: A new dataset of air pollutant Particulate Matter (PM) collected using IoT devices on public buses in Delhi. ML problems of "spatio-temporal interpolation" and "anomaly detection in IoT networks" detailed using the dataset.
Abstract: Air pollution is one of the biggest concerns faced by developing countries like India and the world at large. The capital of India, Delhi and the National Capital Region (NCR), sees life threatening air pollution levels. This paper presents a new Particulate Matter (PM) dataset for Delhi-NCR, which contains PM data recorded over three months from November 2020 to January 2021 over an area spanning 559 square Kms. The data has been collected using vehicle-mounted mobile sensors in collaboration with the Delhi Integrated Multi-Modal Transit System (DIMTS) buses. The 13 bus dataset has been compared with the data over the same period obtained from the pre-existing static sensors, which the buses pass by. Several Machine Learning (ML) problems have been outlined, that can be studied using this dataset, two of which, spatio-temporal interpolation and anomaly detection in IoT networks are detailed in this paper. The dataset is public at https://www.cse.iitd.ac.in/pollutiondata, along with appropriate documentation. We will keep augmenting the website as new data get collected, with more buses and other pollutant sensors (SOx, NOx, COx) added to our deployment in future.
Supplementary Material: zip
URL: https://www.cse.iitd.ac.in/pollutiondata/
15 Replies

Loading