Project Detail: Public Lighting Fault Detection System

• Hourly data read from public area (street, park, etc.) lighting meters were processed with Apache Spark, and an automatic work order was opened to field teams for luminaires that did not light up at night but were lit in vain during the day. An average of 120 faults are detected daily.
• Completely open source software is used.
• The use of Excel lists has been terminated, manual processes have been automated, and the workload in these processes has been significantly reduced.
• Failure detection and repair times have decreased, and material damage and accrual loss resulting from repair delays have been reduced.
• Since the malfunctions were resolved before the customer notices, the requests to the call center decreased and customer satisfaction increased.
• Illegal uses of public lighting have been identified.
• License costs have decreased.
• Unnecessary consumption is prevented and energy savings are achieved.
• Active consultancy for the project continues.