Wireless Sensor Networks (WSNs) provide real-time monitoring and management of industrial conveyor systems. As these networks are often susceptible to risks like black hole attacks, where hostile nodes start to consume and discard data, progressively impairing communication and performance, it becomes increasingly important to provide dependable data delivery. This paper presents a Bond graph(BG) approach for detection of anomaly. This method continually traces the change in data flow interactions by dynamically modeling the WSN as BG, providing the accurate detection of anomalies that highlight the path of the attack. As the attack unfolds, the BG structure provides the path for propagation of the attack, resulting in early detection and localization of compromised nodes. In addition, this method traces small changes in network energy consumption over time, highlighting the increasing effect of the attack on the system performance. By providing proactive response and detection, this temporal causal approach enhances the resilience and security of Industrial WSN.