01whole.pdf (16.25 MB)
Adaptable robotic grasping
thesisposted on 2022-03-28, 15:08 authored by Brendan Menzies
Automation is being integrated into a wider range of tasks every year as technology increases to allow its development and growth. A machine that is able to capably perform more than one task or be able to adapt to unknown circumstances is more beneficial than a machine that cannot. This project investigates the capabilities of low cost tactile sensors being used to determine different properties of gripped objects to allow for an adaptable grasping system. These properties include shape, material type and contact area. A circuit was developed to excite the sensors and condition the sensor signal so that the information can be ready by an Arduino Uno microcontroller. Alongside the sensor research, a program was made using Matlab to control a five degrees of freedom robotic arm, with a user interface. The program was also used to control the gripper based on the sensor data. The tests performed on the sensors investigated the sensor characteristics, variable force tests, force response test, phase shift tests and object shape identification tests. With the system developed and tested, there was no strong indication that any of the force-sensing resistors or interdigitated capacitive sensors could determine object properties. With some modifications, this system could be used to investigate other types of sensors to determine their capabilities in object determination.