The application of underwater remotely operated vehicles (ROVs) for dam reservoir inspections in Indonesia has been practiced for decades, yet conventional diver-based or standard ROV methods face limitations in deep reservoirs regarding safety, accessibility, and data accuracy. For depths exceeding 100 m, these approaches often fail to deliver reliable information for engineering evaluation and rehabilitation planning of critical hydraulic structures. This paper presents an inspection of a low-level outlet (LLO) within a 125-m-high concrete face rockfill dam (CFRD). As a vital appurtenant structure, the LLO supports dam safety and emergency drawdown. After more than 35 years of operation, the operator deployed a smart underwater ROV equipped with stereo photogrammetry. The methodology involved systematic acquisition of overlapping high-resolution images of the submerged structure to produce a three-dimensional (3D) representation of the LLO. The results include a high-accuracy 3D model reflecting the outlet’s actual geometry and condition, enabling detailed identification and analysis of cracks and structural defects that conventional techniques often miss. This study demonstrates that smart ROV-based stereo photogrammetry is a reliable and effective solution for deep-water dam inspections, enabling informed rehabilitation decisions and the safe life extension with an installed capacity of approximately 2 GW, serving millions of customers nationwide reliably.