The main focus of this study was to provide a solution to the garbage management problem in the world. As a solution, this study proposed and implemented self-controlled robot application to detect and collect light weighted solid waste within the ground area as a proof of concept.
This robot application consisted four major components namely, robot platform, image processing unit, sensor component and collection arm. Garbage detection algorithm was a sub component of image processing unit which detects the garbage object on the ground area. Then using sensor component, robot reached to the detected garbage by evading the obstacles on its path. Finally, a wide rotating brush in the collection arm, along with conveyor belt and bin attached to the robot platform collect, grab, push, and dump the detected garbage object.
The results of this study have proven that this robot consumed 27 seconds to lift the garbage objects. These evaluated garbage objects are in weight range of 0.5g to 27g. Further, this robot was able to make navigation decisions averagely within 8 seconds, 20 seconds and 80 seconds in the single, double and triple obstacle situations. This concept can be further developed for more complex situations with more accurate garbage classification system such as glass waste, paper waste, food waste and plastic waste.
 |
| Figure 1 : Side view of the Robot |
 |
| Figure 2 : Front view of the Robot |
Supervisors:
Mr. K.A.S.H Kulathilake
Dr. K.G.P.B Jayathilake