Engineering design optimization problems characterized by highly nonlinear, tightly constrained, and mixed-variable formulations remain challenging for conventional metaheuristic approaches, particularly in maintaining feasibility while achieving competitive objective values. The car side impact design problem (CSIDP) is a well-established benchmark that exemplifies these difficulties due to its narrow feasible region and strong coupling between structural and safety constraints. This paper extends the Scuba Diver Optimization Algorithm (SDOA)—a recently introduced population-based metaheuristic inspired by the physiology and behaviour of scuba divers in open water—to the CSIDP. An adaptive depth-regulated framework is designed in which each candidate solution (diver) evolves under an oxygen-guided schedule that controls the transition between global exploration, moderate exploration, local exploitation, fine-tuning, and full reset. The proposed SDOA variant incorporates elite preservation, adaptive mutation, penalty-based constraint handling, and a hybrid local search tailored to the CSIDP. A MATLAB implementation is presented together with a modular flowchart that explicitly connects the algorithmic stages to diver physiology (oxygen decay, depth selection, ascent, and reset). The effectiveness of the proposed algorithm is validated on the CSIDP benchmark through extensive numerical experiments under multiple parameter configurations. The results demonstrate that the proposed SDOA consistently achieves high-quality feasible solutions with improved robustness and competitive objective values compared to several metaheuristic methods reported in the literature. Detailed constraint satisfaction analysis further approves that the obtained solutions lie close to crashworthiness critical safety boundaries, highlighting the engineering relevance of the proposed approach. These findings indicate that the proposed SDOA provides a reliable and effective optimization framework for complex constrained engineering design problems.