Intelligent transportation systems (ITS) rely on automated sensor data from roads/vehicles to provide adequate traffic management solutions. Radio frequency identification (RFID) technology for detecting vehicles with RFID tags is currently being used for data collection. There are many challenges in adopting this technology in mixed traffic conditions, where the traffic stream is composed of vehicles of different categories that move without any lane discipline. The effectiveness of RFID sensors for data collection under mixed traffic conditions in Kerala, India, is assessed in this study. Data from RFID sensors were used to develop real-time prediction models for estimating queues at a signalized intersection for different traffic scenarios. The estimated queues were compared with actual queues, and the results were promising. Thus, the study devised techniques for constructing real-time estimation models that could be applied in intelligent transportation systems (ITS) operating in mixed traffic settings.