Case Study: SOHGA achieves 30% faster parking patrols with Ultralytics YOLO26

Sohga Co cuts parking patrol time by 30% with Ultralytics

Sohga Co faced the operational challenge of efficiently and consistently monitoring vehicles in large parking lots, a routine but demanding task required by industries like healthcare and pachinko in Japan. Their manual process was difficult to standardize and placed a significant concentration burden on staff. To address this, Sohga implemented Ultralytics YOLO26 as the core AI engine for their new MEGURU mobile patrol system.

The solution from Ultralytics involved running two YOLO26 models in sequence on a standard iPhone to first detect license plates and then recognize their characters in real-time. This provided patrol staff with instant audio feedback, distinguishing checked from unchecked vehicles without breaking their visual focus. As a result, Ultralytics enabled Sohga to reduce average patrol time by 30%, with one documented case showing a time reduction from two hours to 40 minutes. The system also improved staff concentration and is now deployed on over 140 devices across more than 100 clients in Japan.

Sohga Co