Using advances in AI to combat illegal timber trade
By Naren Ramakrishnan, Thomas L. Phillips Professor of Engineering and Director, Sanghani Center for AI and Data Analytics, Virginia Tech Innovation Campus
October 31, 2024
When you read headlines about the war in Ukraine, you probably don’t think about the illegal international timber trade.
There are, to be certain, bigger and more universal concerns. But a huge economic story is unfolding under the cover of the invasion – a plotline that has echoes from the Liberian war financed by its own looted forest – that researchers at Virginia Tech are using AI to help fight.
World Forest ID, an international organization aimed at forest preservation, provided initial seed funding that allowed the Sanghani Center for Artificial Intelligence and Data Analytics at Virginia Tech to explore automated methods to detect illegal timber trade. Armed with the results of that research, the university and World Forest ID — along with partners from the University of Washington and Simeone Consulting, LLC — earned funding from the National Science Foundation to explore further.
The world timber landscape
More than 20 percent of the world’s forests are located in Russia, with the international trade in forest products from Russia, Ukraine and Belarus accounting for a quarter of global timber trade in 2021. In February of 2022, Russia invaded Ukraine, prompting the European Commission to institute a ban on direct Russian lumber imports into the EU, which has been enforced since that July. Despite the invasion, Ukraine has continued to produce and export forest products. Meanwhile, Russia’s domestic forest sector, as well as the global community that relied on Russia’s and Belarus’s wood fiber, has been severely impacted by ensuing official sanctions enacted by countries, as well as the fallout from businesses and organizations voluntarily curtailing or suspending operations, such as decisions made by Forest Stewardship Council (FSC) International and FSC Russia.
Before the war, Russia’s FSC-certified forests constituted 26.4 percent of all FSC-certified forests globally. But after the invasion, FSC terminated all sustainable forest management certificates and trading certificates in Russia. The fallout from Russia’s war in Ukraine continues to drastically alter the global trade of wood and forest products, and has direct impacts on forests, forest conservation efforts and illegal timber harvesting and illicit trade. In addition to the direct impacts, there are also indirect effects felt across the world, as shifting forest product markets are changing the demands placed on forested lands globally. They are also changing the incentives for people to manage those lands through responsible and sustainable forest management practices.
How our AI approach works
Despite the bans, Russian timber is still making its way into markets with active sanctions or bans on direct imports from Russia. We know this through Virginia Tech’s AI work, which assists in increasing the accuracy of geospatial models for stable isotope ratios of various elements (such as hydrogen, oxygen and carbon), which are known as isoscapes.
Isoscapes are landscape maps of isotope distribution, which help us connect timber products to particular regions where the wood was harvested. By using AI to model patterns in the elemental makeup of these samples, we construct multi-layer isoscapes, in combination with advanced statistics, to model the probability that a specific physical wood sample taken from a traded forest product — such as furniture — may have originated from and been harvested in an area that is suspected of being either illegally logged, or that may have been illegally traded or in circumvention of sanctions.
Broader impact and applications
This partnership has been a shining example of the way that higher education, government and real-world organizations can cooperate to deliver results that none of the three could achieve independently. The Sanghani Center — headquartered at the Virginia Tech Innovation Campus in Alexandria, Virginia — is bringing together computer scientists, engineers, mathematicians and statisticians to help solve our nation’s most pressing challenges.
Combining chemical data with advanced data modeling allows the team to deploy World Forest ID’s traceability far more cost-effectively and at a much larger scale. For industry, this helps proprietors comply with the Lacey Act and other timber import regulations here in the U.S., while enabling investigators and prosecutors from agencies like the Department of Justice and Customs and Border Protection to verify compliance. And it allows researchers at Virginia Tech to learn how to develop and deploy cutting-edge technology on the front lines of the illicit timber trade.
Our work also has many other practical applications, as stable isotope geographic determination methods can be applied to origin testing for a variety of organic products. Our work is particularly relevant given the EU’s regulation on deforestation-free products and forest-risk commodities (EUDR) which will require regulated operators to prove that products made from forest-risk commodities (e.g., cattle, wood, cocoa, soy, coffee, palm oil and rubber) do not originate from recently deforested land or contribute to forest degradation from December 2024. Additionally, there has been increased scrutiny on the use of forced labor to produce products and food that enter the global supply chain. Products such as garments and apparel made from cotton produced in the Xinjiang region of China by Uyghurs, and fish and seafood harvested by forced labor, have particularly complex supply chains and can have their geographic origin tested using stable isotope ratio analysis.
As our research progresses in addressing the technological challenges occurring behind the scenes, our project exemplifies how machine learning can assist regulators in posing and answering complex questions. Employing scientific methods to determine product origin on a large scale helps identify and counteract prevalent deceptive practices, such as routing products through intermediary countries and falsely declaring their origin.
The Virginia Tech Innovation Campus is slated to open for classes in January 2025. Every day, we seek to expand our network of partners in AI and other areas of research. We place a strong emphasis on addressing the ethical and societal implications of AI by establishing guidelines and best practices for developing AI systems that are transparent, fair and accountable.
The content is paid for and supplied by advertiser. The Washington Post was not involved in the creation of this content.
Content From
