The XyloTron brings wood identification capability to the field.
The novel device will help researchers, customs officials, and other enforcement authorities quickly and accurately identify timber species.
This capacity will speed the processing of wood shipments and detecting of illegally harvested timber on the spot.
Every day, people across the globe use numerous forest products, from hardwood floors to paper notebooks to mushrooms and berries. Rural residents of developing countries are particularly reliant on the natural resources provided by forests for firewood, construction materials, and food. Beyond these direct uses, forests also provide an array of ecosystem services, including carbon sequestration, water filtration, erosion control, and habitat for a diverse array of species.
Disturbingly, humans are jeopardizing the critical forest ecosystem goods and services around the world through illegal logging for timber products and agriculture.
Illegal logging is not only environmentally detrimental; it also results in significant economic losses and social problems. Harvesting, transporting, and processing wood illegally depresses the global price of legal timber. According to the World Bank, the global market loses an estimated US$10 billion per year, and governments lose an additional US$5 billion in revenues due to illegal logging practices. The law enforcement burden falls on customs officials in both exporting and importing countries.
One barrier to effective enforcement is the lack of a mechanism to quickly and accurately identify timber species prohibited from logging under national law or the Convention on the International Trade of Endangered Species (CITES). Identification of illegally logged species currently requires microscopic examination of wood anatomy by trained experts and/or extensive genetic testing in an off-site lab. To bring wood identification capability to the field, the USDA Forest Product Laboratory (FPL) is developing a portable wood structural scanner and complementary identification software – the XyloTron.
The product name sounds like a new Transformer, and while the XyloTron is much smaller in size, it is just as exciting (at least to customs and forest officials). The prototype of the FPL’s machine vision project consists of a field-portable macroscope, the XyloScope, and identification software. The XyloScope has an array of LED lights to illuminate a cross-section of wood and a scientific-grade camera to capture a magnified image.
The image of the wood structure is transmitted via USB to a computer, where the software uses wavelets to extract horizontal, vertical, and diagonal information from the image at 10 spatial scales (see figure below). This produces a spectral signature with 30 points, which is statistically compared to a reference data set. When restricted to the common commercial woods of Central America, the XyloTron can identify genus and likely species, all in a matter of seconds (Hermanson et al. 2013). It can accurately distinguish among species within the genus Swietenia (true mahoganies) up to 80% of the time. In comparison, trained experts can’t identify among the species with a hand lens, or even a microscope (Averil 2014).
The XyloTron has not been used to identify a specimen’s geographic origin. Identifying the wood’s source is important because specific species may be legally harvested from some areas, but not others. Illegal loggers often falsify timber origin in import/export documentation, especially when they have harvested from a protected area. Determining geographic origin of timber is possible, but it requires DNA-based or stable isotope methods and a wide array of reference material (e.g. a database of DNA data with relevant geospatial variability, likely including leaf samples from tree stands from within and outside protected areas).
The fight against illegal deforestation has reached a critical point – while forests from the Amazon to Indonesia are rapidly diminishing, forest authorities and tech developers are rising to the challenge. In Brazil, the State of São Paulo, Amigo da Amazonia (SPAA), a collaboration between São Paulo state’s Environmental Police and the Department of the Environment’s Forestry Institute, employed wood anatomy experts and trained law enforcement officers at hundreds of checkpoints. With their combined expertise and effort, from 2010 to 2014, police imposed more than BRL $10 million (and up to BRL $27 million) in fines per year.
FPL, in collaboration with SPAA, will be testing and improving the XyloTron to complement and facilitate identifications by wood experts to further crack down on the trade in illegal Amazonian timber. The FPL team is working to finalize an identification model for commercial Brazilian woods that will be field tested in Brazil by the SPAA program. FPL also collaborates with labs around the world to test the XyloTron and to add wood images to the reference library, increasing the breadth of species the system should be able to identify correctly.
Field deployments by expert John Hermanson at the Port of New Orleans and Seattle-Tacoma International Airport suggest that the XyloTron, when used correctly, is more accurate than a field officer after a week of training (Wiedenhoeft and Hermanson 2015).
FPL is currently focusing its model development on Central American and Brazilian tree species, two major sources of illegally harvested wood. While the XyloTron is not yet commercially available, FPL lists its cost at about US $2,000 per unit. Its use by regional customs agents, police, and other timber trade officials should help slow the movement of illegal timber across state boundaries.
Here Alex Wiedenhoeft of FPL demonstrates the XyloTron’s ability to distinguish the genus of wood samples and explains the device’s importance:
Video credit: Andrew Averill, Hardwood Floors Magazine.
Acknowledgement: We thank Alex Wiedenhoeft for his helpful explanations and comments on this post.
Hermanson, J., A. Wiedenhoeft, and S. Gardner. 2013. A machine vision system for automated field–level wood identification. Global Timber Tracking Network. Presentation at Regional Workshop for Asia, Pacific and Oceania on identification of timber species and origins, Beijing, China.
Wiedenhoeft, A. and J. Hermanson. 2015. What does it take to scale up the XyloTron? Forest Legality Alliance, 12th Semi-annual Membership Meeting, Washington, DC.
Averill, A. 2014. Wood ID in Two Seconds. Hardwood Floors Magazine 27: 4.
Gardner, S. Forensic wood science: Science & technology for compliance and enforcement. U.S. Service, International Programs. Power Point presentation.