Artificial Intelligence (AI) Manhole Management Program

manhole and streetThe Costa Mesa Sanitary District (CMSD) uses artificial intelligence and machine learning to detect sewer manholes, rate their condition, and identify those that need repair.

CMSD has approximately 218 street miles of sewer pipeline and about 5,000 manholes within its service area, which makes identifying and determining the condition of every single manhole both time-consuming and dangerous! In the past, this process was done manually by CMSD’s wastewater maintenance crews. But in 2020, CMSD reached out to SpringML, a data solutions company that developed a machine learning model to help the City of Memphis detect potholes. SpringML determined that the same process could be used to identify and analyze CMSD’s manholes.

Here's how it works: a GoPro camera is attached to the hood of CMSD's Code Enforcement vehicle. For two weeks during each quarter, the Code Enforcement vehicle uses the GoPro to record videos of the street during its regular routes. The videos are then processed and analyzed by the machine learning model to identify and rate the condition of each detected manhole. Several months were spent developing the model and then teaching it to accurately rate the condition of CMSD’s manholes with a 1-5 grading system (1 being the best condition and 5 being the worst condition). This innovative technology has increased the frequency of CMSD’s manhole analysis from annually to quarterly, making the process of replacing damaged manholes more proactive and efficient!