Abstract

Diesel trucks are a major source of pollutants, including particulate matter (PM) and NO$_x$. Roadside emissions monitoring can identify high-emitting diesel vehicles, but conventional sensing systems rely on expensive, complex sensing instruments that limit widespread deployment. Low-cost sensors offer increased scalability due to their small size, low cost, and low power consumption, but they have lower accuracy and slower response time. This thesis presents the design and evaluation of a low-cost roadside emissions monitoring system, establishing the foundation for a scalable and widely deployable alternative. We apply machine learning calibration to roadside emission monitoring to improve the usability of low-cost sensors in such systems. Low-cost and reference-grade (high-cost) sensors are co-located and evaluated in both laboratory and real-world roadside environments. This requires the development of a reliable, low-cost data-collection infrastructure, including sensor-interfacing hardware, embedded software, and a server for data storage, visualization, and processing. The resulting system reliably collects data of sufficient quality for emissions analysis and machine learning applications. Emission rates are computed from diesel exhaust plume peaks and used to compare the performance of low-cost and high-cost sensors. To mitigate known limitations of low-cost sensors, we use a long short-term memory (LSTM) model on plume-centered data windows to calibrate sensor outputs. In laboratory experiments, LSTM calibration substantially improves agreement between low-cost and reference measurements. In field deployment, the model improves emission-rate estimates but does not improve high-emitter identification. Overall, this work demonstrates that machine learning can significantly enhance the performance of low-cost air-quality sensors for roadside diesel emissions monitoring. While not fully replacing high-cost systems, the proposed approach narrows the performance gap and supports the development of cost-effective, scalable monitoring networks.

Degree

College and Department

Ira A. Fulton College of Engineering; Electrical and Computer Engineering

Rights

https://lib.byu.edu/about/copyright/