Opportunity to integrate machine management data, soil, terrain and climatic variables to estimate tree harvester and forwarder performance

Abstract

The Cut-to-Length (CTL) harvesting system is nowadays predominant in the field of mechanized forest operations, consisting mainly in harvesters and forwarders forestry machines. These machines are equipped with an On-Board Computer (OBC) that collects a large amount of information concerning machine parameters such as harvested timber, travelled distance or fuel consumption. Stream machine data are sent to the machine fleet management system (FMS) on cloud, stored and automatically summarized on hourly, daily, weekly, or monthly basis. Understanding the benefits of data mining techniques - in finding trends and patterns - exploiting FMS database in relation to topographic and climatic condition is still an ongoing open research question. The present work aims at verifying if and how machine´s performance indicators (e.g. fuel consumption) recorded and summarized on a hourly basis by the FMS are influenced by site specific parameters, such as terrain morphology, soil type, wet soil condition, and weather conditions, derived from open source portal. A specific methodology in machine data acquisition and datasets implementation has been set in this study. The dataset results in a combination of three sub-datasets, consequently merged, filtered and analyzed. A first sub-dataset is made up of “machine data”, a second is made up of “environmental data”, and a third set of data is made of “climatic data”. The obtained results revealed that the combination of different data sources’ provides significant insight into understanding machine performance. Moreover, the integration of terrain morphology and climatic data have direct impact on the machine fuel consumption, harvester machine in particular. However, in order to address specific interactions among variables with greater robustness, further investigations into this project will consider the whole set of variables on a smaller scale (e.g., case study) with higher data resolution.