Decision at individual-log accuracy

Jori Uusitalo, Professor of Forest Operations and Logistics at the University of Helsinki.

Jori Uusitalo, Professor of Forest Operations and Logistics at the University of Helsinki, Finland.

Precision harvesting is the future of logging. In this method, forest management is based on a predefined virtual model and the operations are planned tree-by-tree, while respecting environmental values.

Forest management in Finland has long been based on planning at the plot and stand level. The underlying premise has been that the forest type and the development class of the stand within an individual plot are similar, so the chosen processing method can also be the same throughout the plot. “This, however, is not always the case because within the plots there may be variation that should be taken into consideration in forest management,” says Jori Uusitalo, Professor of Forest Operations and Logistics at the University of Helsinki, Finland.

In fact, Uusitalo predicts that, where possible, a new multi-targeted harvesting model may eventually emerge alongside plot-level planning. Uusitalo calls this new model precision harvesting. Uusitalo has built the model in collaboration with his research team. At the core of the model is the ability to see all the trees in the forest and to plan and carry out harvesting with spatially realistic geoinformation. “A virtual model of the forest is created and then used to plan harvesting at the individual tree level, while respecting environmental values,” Uusitalo says.    

A model based on trafficability, tree stand inventory and logging trail network 

The virtual model of the forest consists of a trafficability model, tree stand inventory and logging trail network, which are combined into the most precise set of data possible. Trafficability describes the load-bearing capacity of the soil, which is impacted by the soil type and its moisture content. The calculations require modeling of the groundwater based on topographical data. Getting an accurate picture requires Uusitalo’s research team to collect data from multiple different sources. 

“Predicting trafficability is challenging. Comparatively speaking, it is easier to predict tree stand data.” The source data used in the tree stand inventory is from airborne laser scanning. In the tree map, the tree stand location data is combined with the key stem indicators, like tree species, diameter, length and possibly also quality factors. “High-density laser scanning started in 2020 and provides ten times more accurate point cloud clustering compared to its predecessor. Utilizing this data is interesting and offers opportunities.”

According to Uusitalo, optimization of the logging trail network is one of the most interesting challenges in forest technology research. “Optimization of the routes should be based on data about the soil attributes as well as on data about the location of old logging trails. In January, our research team published a study in which the old logging trails that are difficult to detect can be extracted using the high-density laser scanning.”

Getting the most out of the wood harvested with precision tree harvesting

The finished virtual model has multiple uses. Prior to logging, the forest owner or logger can, e.g., test the impact of different forest management methods on the structure of the remaining stand of trees, on the removal of the stand of trees, and on the forest’s ecological values. The buyer, on the other hand, may be interested in how to maximize the value of the stand and on minimizing the costs and emissions of transporting by simulating tree stand inventory and merchandising. 

Uusitalo sees that precision harvesting is a response to the divergent and increasing expectations that are currently being put on forests and the forest industry. Even though the forest stock has long been on an upward trend in Finland, merely maintaining the forest stock is no longer sufficient to demonstrate that forestry is on sustainable footing. Sustainable forestry also takes into consideration the requirements of species, recreational use, and the role of forests in regulating the atmospheric carbon balance.    

It is also increasingly important that wood harvested from forests are economically utilized to the widest scale possible. End uses include not only lumber or pulp, but also, for example, highly refined bio-products, textiles, or raw materials for the pharmaceutical and chemical industries.

Automation and robotics are coming to the forest

Uusitalo notes that precision harvesting is also an opening for the world of automation and robotics. “The forest is one of the most challenging environments from an automation perspective. Forest machine movements are multidimensional, and using satellite data to locate a machine in a tree-canopied environment is difficult. Additionally, the forest machine should take into account changes in the soil’s load-bearing capacity and possible precipices.”  

Despite the challenges, work is ongoing and a planning tool is being developed. Uusitalo believes that the next ten years will bring significant advancements in forest machine automation. “But it’ll probably be 20-30 years before we see a completely automated forest machine suitable for commercial use.”

Text: Maria Latokartano / Image: Laura Vesa / In The Forest 1/2022

Precision harvesting solutions by John Deere Forestry introduce to the market advanced technology which assists forest machine companies and operators towards productive and efficient, sustainable forestry. TimberMaticTM Maps and TimberManagerTM expand on the core technology features in each machine to unleash a powerful new set of possibilities. They streamline operations at the work site and the back office.