Optimising rock fragmentation is critical for efficient load and haul operations in hard rock mining. Poor fragmentation leads to reduced shovel productivity, increased energy consumption, and higher maintenance costs. This study investigates the relationship between particle size distribution and shovel bucket payload using AI-enabled machine vision and load sensors installed on a HITACHI EX3600B backhoe shovel. The goal is to identify the optimal P80 (80 per cent passing size) that maximises shovel performance. To meet this goal, an important parameter monitored after each blast was the uniformity index. The uniformity index is commonly used as an indicator of fragmentation distribution, describing how uniformly particle sizes are distributed within the blasted rock pile and the extent of size variability.