The document summarizes a data science capstone project that aimed to predict Metro reliability after Washington D.C.'s year-long SafeTrack maintenance project. The team analyzed historical disruption and ridership data to simulate different post-SafeTrack scenarios. Their results indicated that SafeTrack repairs must reduce disruption severity and frequency by 30-50% for riders to experience improved trip safety and reliability. The project also developed a software tool to simulate the Metro system and visualize the results of different scenarios.