Municipal Waste Characterization and Weight Estimation

---

Objective: Conduct a feasibility study for a neural network that is not only capable of classifying municipal recyclables but also capable of inferring the mass of the object detected for auditing purposes.

Solution: Deploy IntelSight to Municipal Recycling Facilities (MRFs) to gather data and train a segmentation convolutional neural network. Upon completion of the data acquisition and neural network training, the developed neural network will be evaluated and studied.

Status: IntelSight deployed into Material Recovery Facility and captured images and weight data for a variety of municipal waste.

Future: Test neural network on new material to judge accuracy.