Traditional approaches to disaster management in collapsed infrastructure rely on manual search and rescue operations, which are time-consuming, inefficient, and risky for the safety of first responders. The absence of real-time data and situational awareness further complicates accurate decision-making and resource distribution. IoT technology provides an opportunity to revolutionize disaster management by employing smart building management systems for monitoring and early warning. This paper studies the existing targeted Search and Rescue (SAR) operation frameworks and proposes a framework for SAR called Targeted Rescue Operations (TROPS), consisting of Infrastructure Management Systems (IMS) and Mobile Rescue Assistant Platforms (MoRAP) for rescuing victims based on the infrastructure criticality score (ICS). The ICS considers occupant count, habitability, and structural integrity. We implement an IoT-based architecture to collect the sensor data. To estimate the number of occupants, machine learning algorithms are applied to sensor data for passive detection. This approach aims to optimize search and rescue operations and thereby reduce rescue time.