American Water Resources Association
Utilizing Machine Learning to Explore Trends in GSI Inspection Data for Enhanced Sustainable Operations
Presented by:
Achira Amur
Villanova University
ABSTRACT:
Green stormwater infrastructure (GSI) is a strategy used to manage stormwater and offset the challenges facing urban development on stormwater. While these systems are highly impactful, they also require sustained maintenance, which can be costly. Data driven maintenance can create targeted maintenance protocols to reduce this cost. However, municipal stormwater programs are limited by a lack of integrated data and associated analytics to quantify GSI performance and expected life-cycle dynamics. Using data across disciplines, an in-depth analysis can be conducted to understand the performance of GSI holistically and the spatial factors that impact performance over time. Machine learning (ML) is a possible solution to move past barriers in data integration and information extraction. This research uses a variety of ML models to understand features impacting GSI performance, which can inform GSI inspection to ultimately reduce maintenance and improve GSI sustainability. The results from the ML analysis show that the models are able to predict the inspection ratings. This can highlight the parameters that have a substantial impact on the overall rating of the system, allowing for preventative maintenance to ensure long-term functionality. The outcome of this project will be rapid gains in GSI sustainability knowledge that will strengthen core scientific understanding of GSI processes and urban environments, which will be widely applicable across academia and private and public industry.
SPEAKER BIO:
Achira Amur is a Postdoctoral Scholar at Villanova University, specializing in green infrastructure systems and Nature-based solutions, with a strong focus on advancing sustainability through collaborative partnerships across academia, public, and private sectors. A key aspect of her work involves assessing maintenance and inspection trends for green stormwater infrastructure systems. Through spatial and temporal analysis, she and her team have developed tools that facilitate the implementation of proactive maintenance practices, utilizing machine learning to predict and optimize performance. During her PhD, Achira analyzed the performance of a rain garden over a span of fifteen years, using onsite rainfall data to assess rainfall patterns and the garden's response in comparison to regional design criteria. This research included evaluating rainfall distribution, estimating the magnitude and frequency of events, and tracking temporal changes in the site's performance. Prior to her time at Villanova University, she earned both her undergraduate and graduate degrees in environmental engineering from Drexel University.
Engineers:
This seminar does qualify for 1.0 Professional Development Hour (PDH). A Certificate of Attendance will be available for AWRA-PMAS members only. The meeting price for non-members who wish to receive a Certificate of Attendance for the PDH is $10.00 ($3.00 for meeting + $7.00 for certificate).
Please note:
Thank you!