Poster: Toward robust, fast, scalable autonomous plastics sorting

Plastic sorting provides economic, environmental, and energy benefits. Autonomous plastic-sorting systems must be robust, affordable, and high-throughput to efficiently process the large volume of plastic waste generated across the US. Existing approaches, including RGB, IR, and thermal sensing, involve trade-offs among these requirements. In this poster, we present ongoing work on an accurate, robust, and high-throughput sensing system that combines a multispectral IR camera, which reduces the cost of traditional reflectance-based hyperspectral IR systems, with a thermal camera that tracks samples on a moving conveyor belt, thereby reducing the measurement time of Thermal Imaging-Based Plastic Sorting (TIPS).