Providing an ML solution for the turbine audibility problem to Aercoustics Ltd.

Introduction

This project presents a unique opportunity for developers to contribute to environmental sustainability while gaining valuable experience in machine learning for audio applications.

Wind energy in Canada accounts for 3.5 percent of electricity generation and is the second most important renewable energy source in the region. However, during the commissioning of a new wind turbine generation facility in Ontario, provincial regulations require that a study be performed to assess the noise pollution impact on surrounding residents. The study involves collecting thousands of sound level datapoints over the course of months, using these to establish sound level as a function of wind speed with turbine operation and classify datapoints on audibility. The volume of data required to be collected under various weather and operating conditions, in addition to stringent data quality requirements, means that the measurement campaign is long and subsequent analysis is very rigorous and costly.

To automate this process, UTMIST will be working with the sound consulting company Aercoustics to provide an efficient ML solution. The timeline of this project is expected to span across the Winter and Summer term. We are recruiting both a director and developers who can commit at least 6 months to this project. You will be expected to design the ML implementation plans based on Aercoustics’ business demands, deliver a minimum viable product (with a command line interface), and meet their regular KPIs.