This paper presents the key engineering and operational lessons learned from the implementation of autonomous haulage systems (AHS) at Newmont Boddington, with a focus on two practical case studies. The intent of the paper is to share real, site-based learnings that highlight how short-term planning, design adaptation, and operational discipline are critical to achieving safe and productive autonomous operations in a deep open pit mine. The first case study focuses on the use of widened drop cuts and modified spur designs to support continuous mining during bench turnover into the next reduced level. At Boddington, drop cut development is a major bottleneck due to limited space at the bottom of deep, spatially constrained pits. Conventional drop cut designs proved insufficient under AHS conditions, leading to congestion, controller interventions, and reduced productivity. By increasing drop cut width and incorporating engineered spur access, the operation was able to maintain shovel productivity, improve truck flow, and safely manage bench turnover activities without stopping production. These short-term design changes proved effective in de-bottlenecking critical areas and improving overall system performance. The second case study examines the application of spine roads on an oxide waste dump operating under AHS. It demonstrates how 50 m wide spine roads, 110-degree intersections, dump optionality, and disciplined planning practices allowed autonomous haulage to operate reliably through the wettest part of the year. The spine road set-up and winter operating strategy implemented on the 11WD (11 waste dump) in 2025 significantly reduced the need to divert trucks to alternate dumps as well as controller intervention to recover if stuck. As a result, cycle time variability was reduced and the gap between planned and actual performance was minimised. Together, these case studies show that AHS success relies on thoughtful design, operational flexibility, and proactive planning rather than technology alone.