Development and evaluation of an open-source tool for machine scheduling in forest operational planning

Forest operational planning supports sustainable forest management by coordinating harvesting activities under tight logistical constraints. Over the past two decades, numerous operations research models have been developed to assist with harvest planning across regions and operational contexts. However, many of these tools are difficult to reproduce, adapt, or implement in practice. This thesis examines this disconnect and evaluates an open-source modelling framework designed to improve transparency, adaptability, and practical relevance in forest operational planning. A systematic review of 23 peer-reviewed studies published between 2005 and 2024 was conducted using PRISMA protocols. The review found that while existing models demonstrate strong technical sophistication—most commonly through mixed-integer programming and spatial decision-support systems—they are frequently hard-coded to specific sites, rarely open-source, and often difficult to modify. These characteristics limit reuse and slow knowledge transfer between research and practice. Based on these findings, design principles and a practitioner-oriented checklist are proposed to support more modular and transparent tool development. Building on this foundation, the Forest Harvesting Operations Planning System (FHOPS) is presented and evaluated as a candidate open-source machine scheduling framework. Using semi-synthetic datasets representative of British Columbia forest operations, rolling-horizon performance is assessed under deterministic conditions to isolate the structural effects of planning horizon length and re-optimization frequency. Results show that re-optimization frequency has a stronger influence on objective performance and schedule stability than planning horizon length. Moderate planning horizons (4 to 8 weeks) capture most attainable performance gains, while longer horizons increase computational burden with diminishing returns. Overall, this work advances reproducible modelling practices in forest operations and provides practical guidance for configuring rolling-horizon scheduling in practice.