Design and Construction of a Black Soldier Fly Larvae Counting Machine

This thesis addresses the need for reliable preprocessing and counting of Black Soldier Fly larvae (Hermetia illucens) to support process control in large-scale organic waste treatment facilities, where manual separation and counting are labor-intensive, error-prone, and unsuitable for scaling. A modular Black Soldier Fly Larvae Counting Machine was designed and built, combining sequential sieving, sedimentation, and machine-vision-based counting implemented on a Raspberry Pi 5 with a Pi Camera. Each module was experimentally tested with five-day-old larvae batches of 10-100 g, evaluating separation efficiency, processing time, counting accuracy, and throughput. Sieving reached roughly 25 g/min throughput, sedimentation removed fine frass at 6.7-14.3 g/min, and the vision-based counting module achieved 96.7-100% accuracy at about 30 larvae/min, establishing a low-cost, modular proof of concept for automated larvae processing.