The correct capture of forest operations information carried out in forest plantations can help in the management of mechanized harvesting timber. Proper management must be able to dimension resources and tools necessary for the fulfillment of operations and helping in strategic, tactical, and operational planning. In order to facilitate the decision making of forest managers, this work aimed to analyze the performance of machine learning algorithms in estimating the productivity of timber harvesters. As predictors of productivity, we used the availability of hours of machine use, individual mean volumes of trees, and terrain slopes. The dataset was composed of 144,973 records, carried out over a period of 28 months. We tested the predictive performance of 24 machine learning algorithms in default mode. In addition, we tested the performance of blending and stacking joint learning methods. We evaluated the model’s fit using the root mean squared error, mean absolute error, mean absolute percentage error, and determination coefficient. After cleaning the initial database, we used only 1.12% to build the model. Learning by blending ensemble stood out with a determination coefficient of 0.71 and a mean absolute percentage error of 15%. From the use of data from machine learning algorithms, it became possible to predict the productivity of timber harvesters. Testing a variety of machine learning algorithms with different dynamics contributed to the machine learning technique that helped us reach our goal: maximizing the model’s performance by conducting experimentation.

In mechanized harvesting of wood operations, in a cut-to-length system, occupational noise is emitted by self-propelled forest machines, which compromises the safety and health of operators. Therefore, the occupational noise levels emitted by self-propelled forestry machines, in a cut-to-length system, were investigated to determine which are in line with current Brazilian legislation. The noise levels issued by 22 self-propelled forestry machines in the mechanized harvesting of wood operations, in Eucalyptus and Pinus planted forests, were collected during a full day of measurement. Taking into account the operations performed and the type of planted forest, homogeneous groups of operators were formed. Based on Regulatory Norms N.9 and N.15 adopted for labor purposes in Brazil, occupational noise levels were analyzed. In relation to harvester operators, 36.4% were exposed to values above the exposure limit of 85 dB (A) and 63.6% to the action level of 80 dB (A). Regarding the forwarder operators, 100% were exposed to values that exceeded the action level. For the analyzed conditions, for the cut-to-length system employed in harvesting wood in forest planted with Eucalyptus and Pinus, the occupational noise levels of the self-propelled forest machines are not in line with current Brazilian legislation for labor purposes.

Among the activities that burden capital in the supply chain of forest-based industries, the activity of road transport of wood deserves to be highlighted. Machine learning techniques are applied the knowledge extracted from real data, and support strategies that aim to maximize the resources destined for it. Based on variables inherent to the wood transport activity, we verified whether machine learning models can act as predictors of the volume of wood to be transported and support strategic decision-making. The database came from companies in the pulp and paper segments, which totaled 26,761 data instances. After the data wrangling process, machine learning algorithms were used to build models, which were optimized from the hyperparameter adjustment and selected to compose the blended learning hierarchy. In addition to belonging to different methodological basis, a CatBoost Regressor, Decision Tree Regressor, and K Neighbors Regressor were selected mainly for providing minimal values to errors metrics and maximal values to determination coefficient. The learning by stack stands out, with a coefficient of determination of 0.70 and an average absolute percentage error of 6% in the estimation of the volume of wood to be transported. Based on variables inherent to the wood transport process, we verified that machine learning models can act in the prediction of the volume of wood to be transported and support strategic decision-making.

Investments in mechanized wood harvesting operations are characterized by demanding capital and presenting uncertainties associated with productive and economic factors. The economic analysis based on the Monte Carlo method promotes accurate results by measuring the risks of these investments. Thus, we analyzed whether the mechanized wood harvesting operation consisting of self-propelled forest machines under conditions of uncertainty is economically viable, applying the Monte Carlo method for risk management. With the observance of wood harvesting in a forest planted with Eucalyptus, technical-economic coefficients were obtained for felling, extraction, and wood bundle processing operations. The time horizon was six years, with the application of the investment decision rules net present value, modified internal rate of return, and payback period discounted at the opportunity cost rate by the weighted average cost of capital. The results of the investment decision rules indicated the economic viability of the mechanized wood harvesting operation and, based on the stochastic simulation, it was possible to consider the investment as low risk for capital holders. Therefore, the mechanized wood harvesting operation consisting of self-propelled forest machines, under conditions of uncertainty, is economically viable.

Choosing the ideal number of rotations of planted forests under a silvicultural management regime results in uncertainties in the cash flows of forest investment projects. We verified if there is parity in the Eucalyptus wood price modeling through fractional Brownian motion and geometric Brownian motion to incorporate managerial flexibilities into investment projects in planted forests. We use empirical data from three production cycles of forests planted with Eucalyptus grandis × E. urophylla in the projection of discounted cash flows. The Eucalyptus wood price, assumed as uncertainty, was modeled using fractional and geometric Brownian motion. The discrete-time pricing of European options was obtained using the Monte Carlo method. The root mean square error of fractional and geometric Brownian motions was USD 1.4 and USD 2.2, respectively. The real options approach gave the investment projects, with fractional and geometric Brownian motion, an expanded present value of USD 8,157,706 and USD 9,162,202, respectively. Furthermore, in both models, the optimal harvest ages execution was three rotations. Thus, with an indication of overvaluation of 4.9% when assimilating the geometric Brownian motion, there is no parity between stochastic processes, and three production cycles of Eucalyptus planted forests are economically viable.

Physical agent noise can be considered one of the main disturbances that compromise the occupational health of self-propelled forest machine operators. We evaluated whether occupational noise levels emitted by self-propelled forest machines employed in the full tree system are in accordance with both the National Institute for Occupational Safety and Health and ISO 1999:2013 standards, while also proposing mitigating measures aimed at protecting the operators hearing. Seventeen operators, who performed wood harvesting operations in Eucalyptus forests in Brazil, were analyzed. Noise levels were collected in a daily shift of eight hours as recommended by the Acoustics – Determination of occupational noise exposure - Engineering method for full-day measurements (ISO 9612:2009). The standards adopted for the evaluation were the exposure action value of 80 dBA and the exposure limit of 85 dBA based on the National Institute for Occupational Safety and Health – NIOSH and on Acoustics – Estimation of noise-induced hearing loss (ISO 1999:2013) Directive 2003/10/EC. The operators were arranged in homogeneous groups according to the Acoustics recommendation – Determination of occupational noise exposure - Engineering method for full-day measurements (ISO 9612:2009), classified by the operations of felling, skidding of tree bundles and bucking. The results showed that 17 self-propelled forest machines exceeded the exposure action value of 80 dBA, of which 10 machines exceeded the exposure limit of 85 dBA. It was concluded that the levels of occupational noise emitted by self-propelled forest machines used in the full tree system are higher than those recommended by both standards, National Institute for Occupational Safety and Health and ISO 1999:2013. Therefore, the allocation of self-propelled forest machines to homogeneous groups allows inferring mitigation actions that protect operators' hearing. The correct use of hearing protectors during the daily workday provides hearing protection for operators in mechanized wood harvesting. Adoption of actions such as maintenance of cabin seals and mechanical components, breaks for fatigue relief, reduction of daily working hours and rotation of operators in different self-propelled forest machines can mitigate the damage to the occupational health of operators.