Sweden has 23.4 million hectares of productive forest land, primarily managed through even-aged forestry, with 96% of nursery seedlings being Scots pine and Norway spruce. Regeneration efforts are supported by mechanized site preparation techniques, such as mounding and disc trenching. However, the reliance on monocultures increases vulnerability to pests, pathogens, and climate-related stress. Mixed-species regeneration offers significant ecological and economic benefits but introduces operational complexity and higher costs. Despite its labor intensity, manual planting remains the dominant method, especially for small private forest holdings, due to its affordability.

This research seeks to advance sustainable forest regeneration by integrating automation and digital decision-support tools. While commercial mechanized solutions have been tested in Nordic countries since the 1960s, high operational and investment costs, coupled with the relatively low cost of manual planting, have limited the widespread adoption of mechanized planting. Recent autonomous planting initiatives, such as Autoplant and Södra BraSatt, demonstrate the potential to improve safety and efficiency while reducing environmental impact. When paired with intelligent planning tools, these systems facilitate species diversification and enable site-adapted regeneration strategies, offering a promising future for sustainable forestry practices.

A primary contribution of this study is the development of the Digital Precision Planning Tool (DiPPT), which determines the distribution of mixed-species planting based on species-specific site productivity and user-defined density constraints. By using site index maps, DiPPT generates spatially explicit planting maps through either a productivity method or an iterative balancing process tailored to user constraints. The tool allows users to optimize the placement of species adapted for within-site variations, promoting biodiversity and improving forest resilience, and sustaining site productivity.

To advance automation in forest operations, a metaheuristic coverage path planner (CPP) was developed specifically for forest terrain, aiming to minimize travel distances for ground-based vehicles. Using a Genetic Algorithm (GA), the CPP considers roll, pitch, and soil moisture thresholds to ensure paths are navigable safely. It was tested on both synthetic and real terrains, demonstrating its effectiveness to safely plan paths through challenging sites. Tests on steep terrain showed that the genetic algorithm successfully navigated the terrain while minimizing vehicle path length and ensuring operational safety.

Since the CPP with GA took a long time to calculate, an improved CPP named TerraTrail was developed as a faster alternative. TerraTrail was later evaluated alongside another CPP, Pathfinder, for both autonomous planting operations and as a support tool for manual operators. These tools utilized Digital Elevation Models (DEM), Depth-to-Water (DTW) maps, and vehicle kinematics to generate terrain-aware routes, offering coverage and path efficiency comparable to that of manually operated vehicles.

Tests with real-world data from a manually operated planting machine, PlantMax, showed that both Pathfinder and TerraTrail achieved better coverage compared to manual planting systems. On average, TerraTrail achieved up to 18% higher coverage than the manually operated PlantMax, with a similar path distance when working under constrained environmental conditions (e.g., wet zones and steep slopes). Similarly, Pathfinder demonstrated up to 19% improvement in coverage and a 14% reduction in normalized path length compared to PlantMax, highlighting the efficiency of automated coverage path planning.

Together, these advancements tackle key challenges in autonomous forest regeneration, focusing on decision-making and CPP. By integrating digital planning tools, automation, and CPP, this research helps offload forest managers, enabling better on-site decisions and improving biodiversity and forest resilience. This thesis promotes ecological sustainability by supporting the planting of mixed tree species instead of monocultures, increasing coverage for planting vehicles, and minimizing path distances, which in turn supports economic sustainability. Additionally, autonomy allows operators to work remotely, improving safety and reducing full body vibrations, which contributes to social sustainability.