Due to unique conditions pertaining to its location and the long distances between managed forests and processing facilities, the Gaspé region of Québec is faced with a high proportion of pulpwood without any buyers. During a harvest, this material is therefore left in the forest and conventional field inventories are required to monitor the amount and quality of unused material. In a context of labour shortages and low prices for marginal wood products, replacement to these expensive field inventories are being sought. In this project, data (diameter of processed logs and product matrix) automatically collected by single-grip harvesters will be used to develop algorithms to predict the amount of non-utilized woody material left on the ground after mechanized cut-to-length forest operations. Through these algorithms, we anticipate being able to predict the amount of non-utilized merchantable woody material so that the frequency of on-the-ground inventories can be significantly reduced.