Abstract
An innovative approach for the automatic operational monitoring of motor-manual felling activities with chainsaw is here described and discussed. This new system of assessment can be considered as a solution for Precision Forestry (PF) applications and can be employed as a ICT tool for the management of the forest companies. Aim of the proposed system is to manage operational information such as: a) positioning of each felling operation inside the forest, b) measurement of the time spent to carry out the felling, c) estimation of the size of every felled tree, and in the end, d) the analysis of the productivity of the felling operations. This experience is based on the operative principle which considers that the lumberjack, during the actual cutting, drives the chainsaw with the engine at the highest number of rpm. During this action, the generated vibrations reach the maximum amplitude. Some preliminary experiments were conducted during thinning operations in a spruce stand (Picea abies), where a dedicated device composed by a triaxle accelerometer (sampling frequency of 10Hz) and a GNSS module were installed on the filter cover of a professional chainsaw (Husqvarna 560 XP). During the test, for each cut, the duration of the vibration with the maximum amplitude and the section cut area were measured and collected, with the assumption that a strong positive correlation exists between these two parameters. The test objectives were to validate such a correlation and to provide a methodology to estimate the volume of each cut plant. To this aim, once entered the diameter of each plant (derived from the estimated section), the volume for each felled tree was estimated through a specific single entry table commonly used by forestry manager. During two consecutive days of thinning operations, 30 trees were felled and monitored and all the operative parameters through the proposed automatic system were recorded. Simultaneously, a manual time study and diameter measurement using respectively watch and calliper, were performed. 20 of these records were used to create a mathematical model for the volume estimation, while the other 10 were used for its validation. Comparing the results obtained by the automatic time study with those from the manual survey, and with the identification of the felling location, very good correlation values were obtained (R2>0.75). Beside this, a difference lower than 5% resulted by the comparison between the estimation volume carried out through the new approach and the classic one.