NSF-funded partnership with Virginia Tech: Tackling the illegal timber trade with machine learning

NSF-funded partnership with Virginia Tech: Tackling the illegal timber trade with machine learning

NSF-funded partnership with Virginia Tech: Tackling the illegal timber trade with machine learning

World Forest ID has received funding for a collaboration with Naren Ramakrishnan, the Thomas L. Phillips Professor of Engineering and founder of the Sanghani Center for Artificial Intelligence and Data Analytics at Virginia Tech, a university-wide effort that brings together researchers from computer science, statistics, mathematics, and electrical and computer engineering to tackle knowledge discovery problems. 

Funding has been granted by the National Science Foundation (NSF) for a research initiative aimed at enhancing the traceability of traded timber products using advanced machine learning (ML) and data science techniques. Integral to this collaboration are Earth Observation and Isotope specialists from the University of Washington and Simeone Consulting, PLLC.

To determine the geographic origin of timber, physical samples from World Forest ID’s reference collections are analyzed using Stable Isotope Ratio Analysis (SIRA). This technique examines isotopic ratios (e.g., oxygen, hydrogen, nitrogen) that vary predictably across landscapes due to climatic and environmental factors. The resulting chemical data is used to train origin models that can help verify timber harvest locations and scrutinize origin claims.

The funded research project aims to optimize ML techniques for SIRA analytics, maximizing the utility of existing chemical data and improving the accuracy of origin prediction with geospatial methods. Researchers will explore how ML can effectively address data gaps in regions where physical sampling is impractical or impossible, in order to inform World Forest ID’s sampling strategy - identifying areas where confident inferences can be made and those where marginal collections will result in the most new knowledge.

Naren Ramakrishnan emphasizes the advantages of combining chemical data with advanced data modeling: “To enforce timber regulations and international frameworks, there is a need for accurate, cost-effective, and high-throughput tools that can be used to identify and trace illegally sourced timber products.” 

This cost-effectiveness improves the scalability and availability of World Forest ID’s traceability tool, helping companies comply with timber regulations like the U.S. Lacey Act, while enabling investigators and prosecutors to verify compliance. It enhances World Forest ID’s capability to accurately trace timber and other forest-related products, effectively combating global illegal sourcing. This project therefore underscores the critical role of technology in addressing complex environmental challenges.

category
category

Data

Data

Date
Date

Friday, September 15, 2023

Friday, September 15, 2023

World Forest ID is an international organization aiming to protect our forests with a science-based solution for product verification. Copyright ©2023 WorldForestID. All Rights Reserved.

World Forest ID
1 Thomas Cir NW, Suite 700,
Washington, DC 20005, USA

info@worldforestid.org

World Forest ID is an international organization aiming to protect our forests with a science-based solution for product verification. Copyright ©2023 WorldForestID. All Rights Reserved.

World Forest ID
1 Thomas Cir NW, Suite 700,
Washington, DC 20005, USA

info@worldforestid.org

World Forest ID is an international organization aiming to protect our forests with a science-based solution for product verification. Copyright ©2023 WorldForestID. All Rights Reserved.

World Forest ID
1 Thomas Cir NW, Suite 700,
Washington, DC 20005, USA

info@worldforestid.org