23-13: Computer Vision Assisted Condition Assessment of Concrete Cooling Towers

Maintaining concrete cooling towers over time has long been a challenge for owners, particularly in obtaining meaningful structural condition evaluations to inform an owner on proactive and preventative maintenance that can be done. Despite having a visual inspection process in place, site conditions can be overlooked and the progression of deterioration may not be fully recognized due to difficulty in access. Technology now exists to couple engineering site inspection with computer vision to obtain a more holistic condition assessment of concrete cooling towers. This paper presents a methodology to sense, detect, and quantify structural deterioration in concrete cooling towers using photographic images collected on site and machine learning, more broadly known as artificial intelligence. By analyzing site photos, deterioration can be marked and quantified for review and decision making by engineers. A review of photos from periodic inspections can also be utilized to track the progress of deterioration and providing recommendations for the frequency of maintenance and repair. With this information, an asset management plan can be developed with repair budgets and time frames for implementation of repairs in a more prioritized manner. By combining site inspection and machine learning, the end goal is to provide more meaningful condition assessments and customized recommendations for a future course of action for periodic maintenance and repair.