The Qiushi Discovery Engine uses a dual-layer architecture to conduct autonomous research on a physical optical platform. It employs Meta-Trace memory to manage nonlinear research phases and experimental evidence. This system moves beyond simple workflow assistance to produce nontrivial results. Practitioners can now automate the iterative cycle of hypothesis and physical testing.