In trying to deal more effectively with urban stormwater runoff, recent research has been focused on implementing smart stormwater systems that optimize volume collection and removal performance. Several green infrastructure technologies on the Villanova University campus have been recently retrofitted to include real-time control to achieve this improved volume removal goal. These systems utilize the physical processes present in green infrastructure (infiltration, evapotranspiration, etc.) tomanipulate the hydrologic timing to optimize performance through actively controlled inflows and discharges from the system. The control is implemented through a cloud-native platform developed by OptiRTC, a technology firm focused on control of stormwater infrastructure. A modeling study was completed of a real-time controlled green and gray roof system on the Villanova University campus. The study site is comprised of an existing extensive green roof and an adjacent gray roof that has been retrofitted to capture excess runoff from the system. Stored runoff is actively released onto the green roof during dry periods to fully utilize the evapotranspiration potential of the green roof while increasing total capture from the system. The system as a whole demonstrates how forecast information, remote sensing, and self-learning components can be integrated into a green infrastructure system to optimize performance. The long-term continuous simulation compared performance of the green roof system to the performance prior to being retrofitted, allowing for the benefit of the real-time control to be evaluated.