Non-recurrent congestion that develops on highways due to traffic incidents represents the major portion of congestion on freeways and costs billions of dollars. Due to the limited resources of land and money to build new or expand existing highways, Traffic Management Centers (TMCs) are employing different sensing systems to quickly detect and clear traffic incidents to relieve congestion. Closed Circuit TVs (CCTV), Police Patrols (PP), wireless phone reports, Automated Vehicle Identification (AVI), and Inductive Loop Detectors (ILD) are different sensing systems that are deployed at varied levels to detect traffic incidents. Currently, investment decisions about these traffic sensing systems are primarily based on allocated budget and sensors costs. This research endeavors to provide a modeling mechanism that is capable of modeling the performance of a combination of sensors to detect traffic incidents. The model, named Logman Model, will help minimize the need for massive real world experimentation. The model, which is based on the Monte Carlo simulation technique supported by Bayesian Inference Theory for sensor fusion, is intended to predict the performance of a group of installed side-by-side sensors in terms of their combined incident Detection Rate (DR), Time-to-Detect (TTD) incidents, and False Alarm Rate (FAR). The model is validated using incidents and traffic data collected from San Antonio and Houston traffic sensor installations. Preliminary results of the model validation have shown promising results. The proposed model could be used as a performance predictor to aid in the decision-making process of traffic sensing system investments. Elements of the research should be extended to performance assessment of integrated sensing systems for traffic state estimation in general.