Harmonic Pattern Discovery Engine: The Key/Lock/Harmony Framework for Systematic Gap Identification in Physical Phenomena

Scientific discovery often depends on exploring the right combinations of physical phenomena, material substrates, and control parameters. However, the space of possible combinations vastly exceeds what human researchers can explore, and systematic blind spots — single-variable experimental bias, disciplinary silos, publication bias against negative results, and cognitive limits in high-dimensional reasoning — leave significant regions unexplored. This paper proposes the Harmonic Pattern Discovery Engine: a machine learning system trained on the corpus of verified physical effects, encoded as structured combinations of phenomena (Keys), materials (Locks), and orchestration parameters (Harmony). The system learns an empirical grammar of successful combinations, then identifies gaps — regions where the grammar predicts high-payoff investigation but no experimental data exists. The approach is validated through retrodiction (identifying historical discoveries before their date), near-miss resurfacing (flagging failed experiments close to successes), and variation ranking on known systems. The system outputs prioritized investigation targets filtered by experimental feasibility, enabling efficient allocation of research attention. This is not a replacement for physical theory or a prediction engine for guaranteed discoveries. It is a research-direction tool that exploits exactly the blind spots that limit human exploration.