“Big Data” already being utilized by forest industry

Although “Big Data” has become a buzz term in business circles in recent years, the forest industry is already well on its way to using Big Data in a number of areas, from machine centres at the sawmill, to woodlands operations.

By Tony Kryzanowski

Technology providers like IBM and Microsoft are making a big deal out of the concept of mining “Big Data”, but it is something many forest companies already deal with—and they are well on the way to exploiting Big Data, especially to enhance their knowledge of their forest resource.

“For me, Big Data is data sets that are too large to be processed with traditional tools,” says Francis Charrette, a researcher in the Modeling and Decision Support Group at FPInnovations, working with such software products as FPDat and Optitek.

“There is a business around Big Data for sure,” he says. “But as a researcher, I can also say that this data can be useful to have a more efficient forestry supply chain.”

In terms of utilizing Big Data at the sawmill, advances have been made over the past five years with software tools to build platforms that integrate data sets from different companies and machine centres.

Francois Leger, a PhD in mechanical engineering and owner of PMP Solutions, says his view is that the concept of working with Big Data is really an effort to try to better predict the future. His Quebec City-based company provides software solutions called manufacturing execution systems (MES) specific to the forest industry. He says that new software tools like Microsoft Power BI are a good example of the type of tools being developed to help companies make sense of Big Data.

“Everyone is looking hard at how they can market that big information,” says Leger, “and my own understanding is that its value is for forecasting.” He adds that the value of Big Data remains to be seen because it still depends on how well developed the eventual mining tools are, the quality of the data being mined, and hiring a qualified analyst to work with the tools.

Robert Larocque, Senior Director for Environment and Labor Market Policies at the Forest Products Association of Canada (FPAC), says the quality and timeliness of the data being mined will be a critical element as to the value of Big Data intelligence gathering, regardless of the quality of tools being developed. He manages economic and mill data bases at FPAC and is familiar with the growing interest in the Big Data concept.

“You need to be sure that the data that you are getting is accurate,” he says. “You don’t want to mine datasets that are not accurate, because it will skew your results.” He adds that one potential roadblock to a company’s ability to tap into high value Big Data sets is confidentiality.

Big Data analysis could prove useful on the lumber marketing front, to make inroads into new markets, as tariffs come down and new markets open up with the potential Canadian approval of the Trans-Pacific Partnership trade agreement as well as a trade agreement with the European Union.

The question that Logging and Sawmilling Journal posed to these experts is whether the Big Data trend is real, and whether the forest industry can benefit from it. Overall, the experts said it is a real advance in technology. The question is how quickly and how deeply forest companies should invest their time and effort to purchase the tools and hire the personnel to take on a comprehensive Big Data mining venture.

“I’m a believer in Big Data, but it’s new,” says Larocque, and is possibly three to five years away from having practical applications in the forest industry.

All those contacted said that the development of tools and training of individuals with the right skillsets—to understand how to properly work with the tools—is very early days. Most of the work around Big Data, especially as it relates to the forest industry, is still taking place in universities.

One area where companies are already making great strides mining Big Data is in resource management.

Alberta-based Silvacom has been at the forefront of responding to the forest industry’s desire for more powerful data gathering and analysis tools since 1983, when hand-held GPS units first arrived on the scene.

Two big data mining and management tools they have developed in concert with the forest industry are called the Comprehensive Automated Land Inventory (CALI) with seven forest companies as collaborators, and the Land-use Social, Environment & Economic (LuSEE) decision support system with 14 forestry partners. This program uses numerous datasets spanning more than 43 million hectares “to enable the evaluation of trade-offs between overlapping values.”

“There’s a lot of pressure on forest companies on their ability to harvest in certain areas,” says Chris Lang, vice-president of business development at Silvacom. “They just want to make sure that good land use decisions are being made using data and science, rather than emotion.” That’s where LuSEE will prove its worth.

CALI takes multiple big data sources like LiDAR and multi-spectral imagery to create a detailed forest inventory. It provides companies with metrics like volume, piece size and tree species, but with more precision, higher resolution and up-to-date information—the value to forest companies being savings in “substantially less field verification”.

Part of the forest industry’s interest in mining resource Big Data is being driven by greater involvement in the bio-economy. Many companies have recognized that it isn’t enough to just have the capability to evaluate a forest resource on an individual tree scale. Companies now want to know the internal and external characteristics of individual trees, such as their density and elasticity, as well as how much biomass, lignin, hemicellulose and cellulose can be generated from harvesting a particular cutblock. This is in addition to their potential for solid wood and chip production.

“We are taking a more in-depth look at working with some of the CALI data now on fibre composition,” explains Lang, “because when you get into producing biomass and using biomass in different feedstock operations, that becomes really important for facility construction. Our work here is leading edge, for sure. It is more than just about volume.”

Forest companies want more detailed information about piece size, log profile, complete biomass, what’s left behind in the cutblock and what could be available for other people to potentially use.

When it comes to product manufacturing, experts have different opinions whether the installation of machine centres, each with its own software tools and data gathering capabilities, can get in the way of working with a single Big Data gathering and analysis platform at the mill level any more. Larocque says that significant advances have been made over the past five years with software tools to build platforms that integrate data sets from different companies and machine centres.

Valley West Controls Ltd. in Langley, B.C., is a perfect example of one data management and analytical software provider that has already heavily invested in providing the forest industry with big processing data management solutions. Company owner, Rob McLean, says “it’s my life’s work.”

He says the company has had a data gathering and management system available to the forest industry for the past 15 years. It is based on a commercially available program called Microsoft SQL Server. It allows companies to capture data from machine centres—manufactured by any significant vendor to the Canadian forest industry—into one database and analysis platform, including PLC data from such leading vendors as Comact, USNR, and Autolog.

“Give me a vendor and we likely have a driver for them to go and grab their data, pull it into a single SQL database, and then be able to create reports based on data coming into that database,” says McLean.

Their clients include West Fraser, Canfor, Millar Western and Tolko, adding that the technology is accessible and affordable to anyone.

“Really what Big Data means to me is just a lot of data that needs to come in faster than what we have been able to do up to this point,” says McLean. “But I don’t know of anybody that is really doing it well yet. We’re dealing with millions of records right now out of the single SQL database, and it’s the difference between three seconds and ten seconds to refresh and pull the data.” McLean says that at present, it seems that providing companies with data capture capabilities at the manufacturing level on a log by log or board by board basis “is more than sufficient for their data needs.”

On the marketing side, FPInnovations’ Francis Charrette says big data analysis, such as analyzing customer buying habits by being able to track where each stick of lumber or sheet of panelboard is used—as well as being able to track individual wood products back to the forest to demonstrate to customers and consumers that a product has been harvested from a certified and sustainable resource—could provide some value to the industry.

The experts agreed that Big Data analysis could also prove useful on the marketing front as tariffs come down and new markets open up with the potential Canadian approval of the Trans-Pacific Partnership trade agreement as well as the Comprehensive Economic and Trade Agreement (CETA) with the European Union. There will be a need to carefully evaluate local wood consumption habits and marketing avenues for Canadian forest companies to make product inroads into some of those new markets.