Characterizing the Settlement of Activated Sludge Based on AI-Assisted Analysis of Moving and Still Images

Morphological and physical characteristics of activated sludge (AS) from 6 different wastewater treatment plants by moving-image and still-image analysis were compared, and important factors for improving the sedimentation of AS and suspended solids in treated water were examined. There were samples in which AS of sufficient size for recognition in the moving image remained above the interface when the interface settled, and the samples in which particles that could not be recognized in the moving image remained. Minute particles above the interface may be removed according to the shape characteristics of AS or otherwise remain in the supernatant and eventually degrade the quality of discharged water. The shape of minute particles that remain in the supernatant and/or the contact opportunity with AS that is larger and easier to sink than the minute particles may be important for the water quality improvement of the effluent. Although it is not yet known whether the behavior of solids in the container actually occurs in the secondary clarifier due to the difference in system size, the use of light-processing and AI-supported image analysis techniques simplifies the introduction to plant and may be used for objective AS characterization without relying on skilled staff.