Supplementary decision system for messages coming from the interpretation of anomalies in time series

The safety of critical structures such as dams, are great concern of authorities around the world. Conventionally, systems based on time series analysis method have been used to detect anomalies in order to ensure the safety in the operation of these structures. This paper proposes a supplementary system able to receive messages and alarms coming from the first assessment performed by the main system with the goal to improve its accuracy and assertiveness. The messages and alarms are emitted by a already deployed software and they are joined at more three categories of data coming from other sources in order to create a knowledge base. From the composition of the knowledge base, the supplementary system performs a new inference and outputs a new message that content the original message with further important details about the monitored structure. The new message enables the engineering team to make decisions more fast and accurate in comparision with the original message. Experimental results from a real application validate the proposed method.