Agnish1234/Guided-Sensor-Placement-for-Air-Quality-Mapping: Guided Sensor Placement for Air Quality Mapping

Read full text Summarise with VIRAT
Authors: Agnish Brahma
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

What is this? This study solves a simple but powerful problem: Where should we put new air quality sensors to get the most useful information? Air pollution is a serious health issue, but sensors are expensive. Placing them randomly or in a uniform grid wastes money and leaves gaps in our knowledge. This program uses a clever algorithm to find the spots that will reduce our uncertainty the most—giving cities and communities a smarter way to monitor the air. How does it work? Learn from existing data – The program starts with a few existing sensors and builds a model of how pollution likely varies across the city. Find the unknowns – It calculates where the model is most uncertain (the places we know the least about). Pick the best spots—Using smart data structures (like an address book for locations and a priority list), it quickly chooses the few new sensor locations that will give the most valuable information. Repeat – After adding new virtual sensors, it refines the model and picks the next best spots. The whole process is tested on a realistic computer‑generated map of pollution, so we can see exactly how well it works. Key results in plain language Our method (greedy selection) reduced prediction errors by 12–15% compared to random placement and by 6–8% compared to a uniform grid. It runs in about 6 seconds per sensor placement on a free Google Colab computer – fast enough to be useful for real planning. The algorithm scales linearly: adding 50 sensors takes only about 30 seconds, proving it's efficient even for larger projects. What's the big deal? Saves money – Governments and organizations can spend less on sensors while getting more accurate pollution maps. Healthier decisions – Better maps mean better warnings for high‑pollution days, smarter urban planning, and targeted clean‑air actions. Accessible science – The code is open, runs on free tools, and can be adapted to other environmental problems like water quality or noise monitoring. Dependencies The project requires these Python libraries: numpy pandas matplotlib scikit‑learn rtree scipy One can install them with the following: pip install numpy pandas matplotlib scikit-learn rtree scipy

Fields of Study & Subjects

Air Quality Monitoring and Forecasting Air Quality and Health Impacts Environmental and Air Quality Management

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
27 March 2026
Language
EN