Shyam's Award-Winning Paper on Equity Challenges in Civil Engineering

We're proud to share that CEE junior Shria Shyam clinched third place at the recent ASCE Mid-Atlantic West Student Symposium Paper Competition hosted by George Mason University.

In her thought-provoking paper titled "Equity in Engineering: The Imperative for Civil Engineers to Assess Machine Learning Models," Shyam explores the growing integration of machine learning (ML) in civil engineering practices while raising concerns about its potential to perpetuate historical biases.

Shyam presents Amazon's attempt to automate resume screening through machine learning (ML) as an example of bias. The model, trained on resumes predominantly from males who had been successful in the past, presented a preference for male candidates, inadvertently depreciating terms linked with women. 

Further, Shyam's exploration of civil engineering's past exposes systemic inequities ingrained in American infrastructure development. From the ill-fated relocation of contaminated soil to Warren County's landfill, disproportionately impacting marginalized communities, to the callous routing of Interstate 40 through Nashville's African American neighborhoods, resulting in profound socio-economic repercussions, the examples underscore the need for ethical recalibration in engineering practices.

Shyam champions a multi-step approach for civil engineers to assess ML models for equity. This method entails meticulous scrutiny of datasets for representation, meticulous examination of model parameters to ensure neutrality, rigorous testing of performance across diverse demographics, and extensive stakeholder consultation pre-deployment.

Through these measures, Shyam asserts that civil engineers can proactively address biases inherent in ML systems, thereby developing a fairer and more inclusive landscape in civil engineering practices.