New Tech Helps N.B. Harvesters See The Forest For The Trees
EDMUNDSTON — The Northern Hardwoods Research Institute is helping the operators of tree harvesters spot the difference between sugar maples, yellow birch, and red oak by combining drone and computer technology.
The Digital Timberlands 20/20 project uses drones, equipped with light detection and ranging (LiDAR) sensors which collect thousands of data points, to accurately describe not only the forest but every tree within it.
“We’re convinced that the future of forestry has to go through digitalization,” says Gaetan Pelletier, the executive director of the Northern Hardwoods Research Institute.
Working with SceneSharp Technologies, a data analytics firm in Fredericton, and with artificial intelligence (AI) students and teachers at the CERFO (Centre d’enseignement et de recherche en foresterie) in Sainte-Foy, Quebec, Digital Timberlands aims to precisely map forest sections in 3D, providing foresters with a highly detailed, digital view of a section of forest so that they can optimally manage it.
The project aims to create a value chain from start to finish, resulting in a tool that will not only help foresters but keep the intellectual property for this emerging technology here in Canada, while exporting it across the globe.
Research opportunity
“We understand where to find the knowledge and talent in our postsecondary institutions needed to solve the problems faced by industry,” says Mitacs CEO John Hepburn. Mitacs has helped partner NHRI with interns to assist with the Digital Timberlands project.
“Especially for small and medium enterprises, finding the talent and expertise that exists in our very strong postsecondary institutions can be a daunting challenge,” Hepburn says.
Bastien Vandendaele is one of the interns brought on board by Mitacs. A bio-engineer specializing in forest and natural areas management, Vandendaele is studying at the University of Sherbrooke. His areas of research include remote sensing applied to the environment, so the opportunity was a perfect fit.
“What I like about this job is its multidisciplinary nature,” he says.
“I don’t just do data analysis, but I go out into the field, deal with service providers, collaborate with foresters, geomatics and computer scientists, transfer knowledge from the academic world to the industry. This is an important aspect of research and development because it is at the frontier of different disciplines that the innovation happens,” he says.
Vandendaele has completed internships in the field in Edmundston and has worked with the members of the 25 industry partners aligned with the project. He also has his sights set on getting his drone pilot’s license soon.
“The most complex part was to understand the entire process chain and to optimize its different components. [Drone] LiDAR technology is relatively new, especially in forestry,” he says. He adds that there wasn’t really a roadmap laid out before him in terms of how to best use the technology at hand.
“Sometimes we had to make our own trial and error until we got the best configuration. Acquiring a 3D point cloud is one thing, but providing quality outputs for the forestry industry is another.”
Economic importance
In 2020, nearly 80,000 hectares of trees, out of approximately 2,900,000 hectares, were harvested on Crown and private lands in New Brunswick, according to the National Forestry Database. According to the Canadian Business Journal, “forestry is the single largest contributor to New Brunswick’s economy from the private sector, which translates into about 3.5 percent of provincial GDP.”
New Brunswick’s forests are partially harvested, meaning some trees are left in place because they are too young, or not the desired type. But many of New Brunswick’s tree harvesters are arriving from countries like Brazil, where the plantations they are harvesting are all the same type, and age, of tree–and where they don’t have to contend with conditions like snowstorms.
That knowledge gap of what to pick and what to leave is becoming an increasingly difficult obstacle.
But what if the machine, the tree harvester itself, could tell the operator which trees are good to cut and which ones aren’t?
“That’s phase two, the actual implementation in the cab of the machine,” says Pelletier.
“We will have to work closely with machine manufacturers to make sure their systems are compatible. We’re hitting all kinds of hurdles. Computers are never as fast as we want them to be and we don’t have a 5G signal in the woods.”
He says that this project is helping to identify and address these challenges now so that when infrastructure like 5G in the woods is ready, they will be too.
“We’re even thinking at some point instead of having to rely on somebody flying a drone with a LiDAR scanner ahead of the operations, that we could have a LiDAR scanner on the harvesting head of the machine. So, we would generate real-time information.”
“But these are all things that we tend to overcome with somebody else’s grey matter,” he jokes.
The data gathered and analyzed by the project will also be an important part of ongoing forest management for not only local, but global threats and challenges.
“We know also these desired hardwood species are threatened due to climate change,” Pelletier added. “If we don’t intervene now and arm foresters with the upstream technology they need to make better forest management predictions and decisions, the entire value chain will be impacted.”
Editor’s note: this story was last updated on November 17, at 5:05 pm.
Alex Graham is a Huddle reporter in Saint John. Send her your feedback and story ideas: [email protected].