Demonstration of automated robotic workcell for hazardous waste characterization

An automated robotic workcell to classify hazardous waste stream items with previously unknown characteristics is designed, tested and demonstrated. The object attributes being quantified are radiation signature, metal content, and object orientation and volume. Multi-sensor information is used to make segregation decisions and to perform automatic grasping of objects. The workcell control program uses an off-line programming system as a server to perform both simulation control as well as actual hardware control of the workcell. The overall workcell layout, sensor specifications, workcell supervisory control, 2-D vision-based automated grasp planning, and object classification algorithms are discussed.<<ETX>>