Detecting Energy Theft through Assignment of AI Detected Buildings to Registered Consumer Units: A Novel Approach

This paper presents a novel approach for detecting energy theft within the electrical network. The method utilizes satellite imagery to identify unauthorized connections by cross-referencing building rooftops with registered consumer units. Efficiently solving the assignment problem between buildings detected by AI and registered consumer units, flagging suspicious locations lacking assigned meters for prioritized inspections. Simulated tests yielded promising results, identifying a majority of illegal connections. Real data applications are presented, discussing subsequent enhancements to increase assertiveness. The method’s practical implications empower utility companies to combat energy theft, improve system efficiency, and ensure the electrical energy sector’s reliability and profitability.