University of Wisconsin–Madison

Ghana Up in Arms: Battling Illegal Timber Trade with the Xylotron

By Aberdeen Leary | October 5, 2018

Examining evidence of illegal logging in a forest reserve, Assin district, Ghana. (Photo by ESPA)

Four research scientists are on campus this semester to learn the ins and outs of UW-Madison wood identification technology in an effort to bring the knowledge back to their home country of Ghana. Emmanuel Ebenyenle, Senior Research Scientist at the CSIR Forestry Research Institute of Ghana, and Ophilous Lambog, Kofi Abban, and Asi Ebeheakey, scientists at the Timber Industry Development sector of the Forestry Commission, are

A comparison between wood types. (John C. Hermanson, U.S. Forest Service, Forest Products Laboratory)

working with Professor Alex Wiedenhoeft at the Center for Wood Anatomy Research in the US Forest Service’s Forest Products Laboratory.

The Center for Wood Anatomy Research team, headed by Wiedenhoft, conduct research on wood, trees, and forests and has curated the world’s largest xylarium, a collection of wood specimens from around the globe. The team is also part of an interdisciplinary cooperation funded by the U.S. Department of State and the Forest Service to develop a “XyloTron.” According to the U.S. Forest Service, the XyloTron is an “open source, field-portable machine vision wood identification system,” essentially meaning a machine capable of determining a type of wood by simply touching it.

The XyloTron is not only a handy tool for scientists, but is essential in helping to combat illegal timber trafficking. Endangered tree species is Ghana are regularly harvested and passed off as common woods. “A few years ago,” Emmanuel Ebenyanle claims, “90% of commercial timbers were [supposedly coming] from just 10-12 species.” That’s where Emmanuel Ebenyanle and his cohort of Ghanaian scientists come in. Ebenyanle originally came to the United States for four years to earn his PhD from Michigan State. Following this, he returned to Ghana, joining the Forestry Research Institute of Ghana. While in this position, a senior colleague at the Timber Industry Development division contacted Ebenyanle to run a training program on wood identification, where his current team would be students. Though Ebenyanle frequently gave these trainings, something in this call prompted him to look for a more high-tech solution. This is when he came across the XyloTron. “So, I contacted Alex Wiedenhoeft,” Ebenyanle says of the moment, “and we put a proposal together to UNIDO.” When the United Nations Industrial Development Organization responded that the proposed project was not feasible, Wiedenhoeft travelled to Ghana, where he and his colleagues studied the country’s wood identification workflow, met with the Forestry Commission, and visited Ebenyanle in his lab to see what equipment the Ghanaian foresters had so far.

The Xylotron uses machine-visioning technology to identify wood species. (John Hermanson, USDA Forest Service)

The team’s second proposal to UNIDO succeeded and the plan was put into motion. Ebenyanle describes the project as twofold. The first requires the scientists to study and improve the XyloTron for the automation of wood identification in Ghana. “If you have a wood and you apply the device to it,” Ebenyanle remarks, “within split seconds, it tells you the name.” With this, a task that previously took these scientists two weeks is complete in under a minute. Ebenyanle goes on to explain that the XyloTron has two parts. In order to master the machine, the scientists must first image the 106 Ghanaian timber samples they brought with them into the XyloTron’s database.

The second phase of the project will serve as a back-up plan, in the event that technology fails while in Ghana. The scientists will develop a Wood Identification Manual of Ghana Commercial Timbers. By making a small cut in the wood and comparing what they see to the manual, timber experts will be able to quickly identify which tree the wood has come from. “So, you have the machine phase and the human phase, and they are complementary,” Ebenyanle boasts.

“When we go back to Ghana, we are going to image the other woods we have over there and share it with the U.S. Forest Product Lab. Then, they’ll add it to increase the data so that they build a robust system for the xylotron,” Ebenyanle says of their efforts to continue the progress toward quick identification of wood specimens. When the manual and XyloTron modifications are complete, the scientists will serve as trainers to timber inspectors across the country.