Fraud detection need in the energy program implementation for Utility companies
By Atul Gunjal, CEO and Ameet Shedge, CTO, iWorkTech
To address energy conservation, generation companies and commercial organizations which deal in services like rating, programs and measure implementations take a multi- prong approach. The nation is cautious about its carbon emissions and aims to earn carbon credits by curbing on these carbon emissions. Thermal energy generation companies, seen today, play important role in this process. The challenge for the company is to meet the energy demand of consumers and controlling emission. One of the approaches taken by the company is to perform an energy audit of the dwelling before and/ or after construction. The audit captures data related to heating, cooling, hot water, lighting, and appliance energy loads and consumptions. The article scope is limited to single and multi-family homes. The company implements an energy conservation objective through the program. This program is mostly implemented by a non-profit organization which focuses on green earth/energy conservation objective. The company allocates a budget and sets a target for implementation. The budget covers operational expenses plus rebate or incentives offered to consumers. The objective of the program gets implemented by the staff and corrective measures are implemented by its auditor or rater. The challenges faced by the CIO is to see that each dollar spend, is well spend and to detect instances of fraud and incompetency. The proactive steps taken towards these challenges help in ensuring that the objective of the program is met.
“This approach ensures that continuous Quality assurance is performed to detect and report any anomalous result in the submitted data”
Program Implementation Platform (PIP)
Data Flow, approach and algorithm
The data flow can be depicted as follows: First, the home data is gathered by the file executive. It is important that this data complies with the energy model. This helps to get consistency in the rating. To achieve this, the following is necessary:
• Clear and consistent information for the rater
• Training needs to cover assessment methods, measures, and data points.
• Independent third-party assessment needs to be done on the field reviews and submitted data files. This is done to detect any fraud or incompetency. o Establish limit on input variables o Determine bounds check o Detect and warn users for input values beyond reasonable limits o Generate XML data which is stored in the repository for pattern detection, future use, data exchange with other systems. This approach ensures that continuous Quality Assurance is performed to detect and report any anomalous result in the submitted data. This continuous approach absolves staff from routine manual checking. The automated nature of the job helps to handle any increase in the volume of homes. This process can also detect existing patterns or establish a new pattern. This pattern is stored in the repository. This rule-based approach helps to build a flexible system. Some of this system’s silent features are:
• Backend system
• Sanitize data received
• Analyze data based on several data validations
• Possess multiple methods for o data analysis o pattern recognition
• Flag files with issues
• Deliver report on analysis
• Unknown nature of data needs to be flagged and further analyzed. This analysis helps determine whether new data needs to be consumed by the system for fraud detection or not.
• Efforts involved to sanitize and manipulate data
• Diverse data models
• Project efforts
• Support needs
• Close communication is needed between involved stake holders
PIP helps CIO to keep focus on the project objective. PIP’s Fraud module helps to get consistent rating and helps detect fraud and incompetency. The fraud detection helps to keep tab on the unfair practices. By taking proactive measures, like training, helps to address issues like incompetency.