Can we use smartphones to measure noise pollution?
Master's thesis at UiO
Topic: Mobile Crowdsourcing for Mapping City Noise
Contact: [email protected]
- Server: https://github.com/papkos/noisemapper-server
- Android app: https://github.com/papkos/noisemapper-app
Mobile phones carried by people are capable of capturing and sharing image, acoustic, and location. We are able to accomplish complex tasks when millions of people can share information and join their effort. This development is effectively leading to a new approach called mobile crowdsourcing. We have seen the applications of crowdsourcing in traffic real-time navigation (e.g., WAZE) and environmental monitoring.
In this thesis, we will mainly focus on mobile crowdsourcing for noise measurement in Norway. In Norway, nearly 30% (about 1.4 million) are exposed to noise levels above 55 dB outside their home and the number is rapidly increasing [1]. In Fornebu, there are recent quick surge of constructions, population, and traffic. Norway has the ambition to reduce 10% noise annoyance by 2020 [2]. However, there is no cost-effective method yet to know noise level in order to provide strategic regulations for government, local community, and residents.
In particular, we will study mobile crowdsourcing techniques, develop an APP in Android system (Google Nexus 6 for experiment), build a city noise map (e.g., Oslo, Fornebu) for PC or mobile phone, and test algorithms performance based on collected data. The APP will conveniently be used by end- users for crowdsourcing based noise level measurements and visualization. It is possible to extend to combine the noise map with other maps of a city (air pollution, property value, incidence of medical problems (psychological and physical problems can be caused by noise pollution)).
Study and develop mobile crowdsourcing for noise measurement in Norway. Develop and test algorithms and methods for energy-efficient background noise measurement. Detect phone location (in pocket, in bag, indoors, outdoors) and adapt the filtering and sampling algorithms to enhance accuracy. Display collected data on a web interface (and also on the device), possibly with other statistical data – e.g. real estate prices, health trends, weather – and try to discover correlations between them.
- study mobile crowdsourcing techniques
- develop an APP in Android system (Google Nexus 6)
- build a city noise map (e.g., Oslo, Fornebu) for PC or mobile phone
- design new algorithms to improve noise level measurement accuracy (e.g., when mobile phones are in pocket), reduce mobile phone power consumption and filter faulty data
- test algorithms performance based on collected data
- extend to combine the noise map with other maps of a city (air pollution, property value incidence of medical problems (psychological and physical problems can be caused by noise pollution))
- Learn techniques on mobile crowdsourcing
- Learn programming in Android system
- Learn how to analyze data and visualize noise data in map
- Learn algorithm design and problem modeling in the context of crowdsourcing data
- Programming experience
- Algorithm design experience
- Yan Zhang
Simula Research Laboratory // IFI, University of Oslo
Email: [email protected] - Stein Gjessing
IFI, University of Oslo
Email: [email protected]
[1] http://www.environment.no/Topics/Noise/
[2] "Quality of the Acoustic Environment", Oslo Community. http://www.miljo.oslo.kommune.no