In this paper, we present a novel approach that can identify operational faults associated with control process by using log data records of the PLC program signals. The proposed approach automatically creates a logical model of the PLC control process, called the unified logical model and then utilizes that model to detect faults. The unified logical model includes the signal-state logical models of the devices and the relational logical models of the device groups. The signal-state logical model is designed to depict the device behavior, and the relational logical model is designed to describe the relationship between the dependent devices. The proposed approach automatically generates the unified logical model from the PLC signal log data records and employs a hash table based model indexing and fault searching scheme to identify the faults in the manufacturing system. Experimental results show that the proposed model can be utilized to detect the faults effectively. © International Journal of Industrial Engineering.