Rapid urbanization has increased the generation of mixed household waste, where wet and dry materials are commonly discarded together. Improper segregation reduces recycling efficiency, contaminates recyclable materials, and accelerates landfill saturation. To address these challenges, this paper presents a Smart Dual-Compartment Waste Separator Bin, a low-cost, sensor-based prototype that performs automatic real-time segregation at the source. The system uses a moisture sensor and ultrasonic sensor integrated with an Arduino microcontroller to classify waste as wet or dry and mechanically direct it into the correct compartment through a servo-controlled flap. A secondary ultrasonic sensor provides real-time fill-level monitoring and triggers alerts as compartments approach capacity. Experimental evaluation was conducted on 200 waste samples, and system performance was assessed using standard evaluation metrics. The prototype achieved 92.4% sorting accuracy, validated through a confusion matrix and accuracy formula, while the fill-level module recorded a mean error of 4.8% across repeated trials. Results confirm that the system operates reliably, responds quickly, and maintains stable performance under varying moisture conditions. The modular design also supports future enhancements, including IoT connectivity for cloud-based monitoring and integration with municipal waste management platforms. Compared with earlier sensor-based smart bins, the proposed system offers improved automation, higher accuracy, and better scalability. Overall, the findings demonstrate that a sensor-driven, real-time monitoring approach can significantly enhance waste segregation efficiency and support sustainable smart-city waste management.