The Brazilian forest sector has been developing every year, whether in planted or conserved areas, with technological applications that have different performance levers in the production chain, from the field to the forest-based industry. In the report of the Brazilian Tree Industry, released in 2021, in Brazil, more than 1 million commercial trees are planted every day, reaching an area of 9.55 million hectares for industrial purposes and another 6 million conservation areas.

This balance has a storage potential of more than 4.5 million tons of carbon dioxide. Given this exponential growth with the expansion of new pulp mills and the entry of new players in the country, the demand for areas with commercial forests to supply and guarantee the production plan of these plants has been growing. In Mato Grosso do Sul alone, there is a projection of an increase in pulp production of more than 6 million tons per year.

With this promising scenario for the entire forestry production chain, silviculture, harvesting and logistics operations are often required to seek operational excellence in their activities and processes, maximizing the production and quality of the wood produced. The world pulp market is increasingly strict with the quality of products in its most diverse applications, from packaging, cosmetics, fabric fibers with nanocellulose to tissue.

In line with this industry requirement, the quality of the raw material must also follow the same criteria in the forest, where the challenges with the impact of wood bark on industrial processes are enormous. They generate high maintenance costs and affect the availability of equipment in chip preparation and paper machines, consumption of chemicals in bleaching and cooking to declassification of batches for dirt and sand content.

This balance between production and quality passes through some pillars that support this balance and depend on monitoring and monitoring of quality to avoid impacts on the production processes of pulp mills. Between each process of the chain activities, evaluation tools can be applied, ranging from a visual analysis of the logs with target lengths, minimum and maximum diameters, tortuosity, debarking quality, mechanical damage, pest attacks, fires, among others.

The results of these evaluations are of paramount importance for directing actions with the development and harvesting teams, aiming at equipment adjustments, operational training, choices of machines and heads most adapted to the application, with a focus on production and quality. As this monitoring stage employs teams that need to be traveling to the harvesting areas, which are highly pulverized and end up generating costs, the sector is undergoing a technological revolution, where automated assessment tools are being developed.

The use of artificial intelligence with the objective of systematizing the process and generating information in real time with traceability has been reducing this investment, increasing the capacity of analysis in the form of a sense of the trucks received at the factories. Currently, the use of camera analysis is being developed and applied through models that were developed with an image bank where artificial intelligence is able to identify in the wood load what is bark and what is log.

This entire process was created in the Digital & Analytics environment, with the collaboration of startups that supported the development of the model built from more than 2,000 images in a “colab” network. Thus, it was possible to generate a better assertiveness of the software with the calibration of manual measurements, aiming at a high correlation of the result of the analysis of Artificial Intelligence in the measurement of the peel index in the load.

This information can be managed by the logistics team inbound for the direction and treatment of this load in the first moment, avoiding that loads with high bark index are consumed. Structurally, information is generated from which location this wood was transported and, consequently, harvested, providing traceability to the development team with operational excellence so that they can carry out the necessary corrective actions. With the application of this tool via proof of concept, it was possible to observe the reduction in the bark content of the chip that feeds the digester, ensuring a management of wood quality in industrial processes, reducing costs and evolving in the links of the forest production chain.