PhD-AECM Doctoral Proposal Presentation: Tannaz Afshar

Title: Decision Support Framework for Residential Deconstruction: Learning from Large-Scale Evidence and Neighborhood-Scale Implementation in Larimer, Pittsburgh

Name: Tannaz Afshar, Ph.D. candidate in Architecture–Engineering–Construction Management (PhD-AECM)

Date: Thursday, April 30, 2026
Time: 10:30am-12:00pm ET
Location: Virtual

Advisory Committee:

Azadeh Sawyer, Ph.D. (Co-Chair)
Assistant Professor in Building Technology
School of Architecture
Carnegie Mellon University

Joshua D. Lee, Ph.D. (Co-Chair)
Associate Teaching Professor
School of Architecture
Carnegie Mellon University

Erica Cochran Hameen, Ph.D., NOMA, Assoc. AIA, LEED AP
Associate Professor
School of Architecture
Carnegie Mellon University

Melissa Bilec, Ph.D. 
George M. and Eva M. Bevier Professor
Department of Civil and Environmental Engineering
University of Pittsburgh

Abstract:
Construction and demolition activities generate over 600 million tons of waste annually in the United States, with demolition accounting for the majority of this material flow. Although deconstruction enables material recovery and reduced environmental impacts, demolition remains the dominant building removal strategy. The shift from traditional demolition practices to deconstruction methodologies is not merely a voluntary industry trend but is increasingly being mandated by legislative action.  Waste management problems are inherently uncertain and probabilistic due to the wide array of possible scenarios generated by the numerous interacting variables and inputs involved in the decision-making process. This uncertainty becomes particularly critical and impactful when examined at a larger scale, such as a neighborhood, where key planning decisions are typically made. However, most existing approaches rely on deterministic or case-specific analyses that limit scenario testing, rapid comparison of alternatives, and integration of both expert knowledge and empirical data. To address this gap, a probabilistic decision-support framework based on Bayesian Networks was proposed to represent relationships between building attributes, contractor experience, salvage performance, and sustainability outcomes.  The framework uses empirical data from residential deconstruction projects to model relationships between building characteristics, contractor practices, and salvage performance. The model enables scenario analysis by estimating the environmental, economic, and social implications of alternative end-of-life strategies. To demonstrate the scalability and real-world applicability of the proposed analytical framework, the methodology is applied to a neighborhood-scale case study in Larimer, Pittsburgh, a context with aged housing stock and ongoing redevelopment. This application provides a practical validation of how the framework can support sustainability-oriented decision-making across an existing building stock.