Meter tampering is defined as a fraudulent manipulation which implies a service that is not billed by a utility company. It is a lack of consumption control for the utility company and a main problem because they represent an important loss of income. We have developed a methodology consists of a set of three algorithms for the detection of meter tampering in the Emasesa Company (a water distribution company in Seville and one of the most important of the country). The algorithms were generated and programmed after a data mining process from the database of the company and they detect three type of consumption patterns: Progressive drops, sudden drops and abnormally low consumption. The methodology has been tested with in situ inspections of the customers of a village of the province of Seville. Once carried out the inspections by the utility, the inspectors confirmed a good success rate taking into account that the detection of this type of fraud is very difficult because it is a noninvasive technique. Besides, this type of detections is a topic that, if we take a look at the state of the art, there are few references or works.

Ministerio de Ciencia y Tecnología TEC2013-40767-R