Abstract
In the global push to achieve digital transformation in operational forestry, researchers and managers face unique challenges. Remote worksites, steep terrain, dense canopy cover, and a lack of connectivity limits real-time data sharing among people and machines, making it difficult implement technologies like teleoperation and automation of logging equipment. This thesis investigates technological innovations to enhance operational efficiency and smart forestry implementation in the Inland Northwest United States. As forestry moves towards more technological, connected, and data-driven systems, understanding limitations to both mechanical innovation and performance and the limits of digital communication in forested environments is critical. This work contributes to the growing need for operational research that integrates equipment productivity with emerging communication technologies.
The first study compares rubber-tired and tracked grapple skidders quantifying cycle times, production rates, and skidding costs using field collected data and modeling. Data were collected during a commercial clearcut timber harvest on the University of Idaho Experimental Forest. Two John Deere 648 grapple skidders were compared in the study, one on rubber tires and with G and R track attachments installed. Observers were positioned at both ends of the skid trail and relayed equipment productive cycle time information via handheld radio. Cycle elements were recorded on a single smartphone, and photos of the grapple were analyzed to calculate total volume per load (MBF). A high accuracy 1-meter resolution GNSS unit was placed on the top of the machine to collect location data and measure skid trail distance. The tracked skidder achieved higher overall production rates than the rubber-tired skidder. Although the tracked machine had slower cycle time, its greater production resulted in skidding costs that were slightly lower but not different statistically; $14.01 per MBF for the tracked skidder versus $14.75 per MBF for the rubber-tired skidder, making the tracked machine $0.74/MBF cheaper. Mixed effects models show the tracked skidder increased volume per turn (p = 0.042) while differences in cycle time and skidding cost were not significant. Overall, the tracked skidder had slower cycle time, modest productivity advantage and equal skidding cost compared to the rubber-tired machine.
The second study evaluated Starlink Low Earth Orbit (LEO) satellite constellation performance under forest canopies, measuring download and upload speeds across different stand conditions. A Starlink satellite receiver was placed at 30 randomized plot centers under a range of canopy densities. Connection time was measured, and speed tests were run thereafter if connection was established. Speed tests were run at the plot center, and 10, 20, 30, 40, and 50 meters from the plot center in each cardinal direction (north, south, east, and west). Upload and Download speeds were recorded (Mbps), and LiDAR was used to calculate stand density index (SDI), leaf area index (LAI), rumple index (RI), and vegetation cover (VC). Linear mixed effects models showed that Starlink upload and download speeds were primarily affected by distance. Logistic regression showed that higher SDI, LAI, and RI reduced the probability of establishing a connection, while VC did not have an effect. Upload speeds were consistently lower than download speeds. Overall, dense canopy cover (indicated by high SDI, LAI, and RI) and distance from the router are the main constraints on LEO satellite connectivity and performance in forested areas.
Results from the studies presented in this thesis highlight potential operational benefits of quad-track skidders that can be achieved at no additional skidding cost and demonstrate the potential of Low Earth Orbit satellite systems to provide reliable connectivity in forested areas. By integrating machine performance data with digital connectivity, this research supports data-driven decision making and advances principles of smart forestry. Together, these findings provide a quantitative foundation for designing more efficient harvesting systems and for determining where and to what extent LEO satellite networks can support future autonomous, semi-autonomous, and digitally assisted forest operations. The combined insights from both studies help identify operational thresholds, technological constraints, and opportunities for future research. These results can help guide forest managers, equipment manufacturers, contractors and researchers as the industry transitions toward smarter, more connected forest operations.