Design and Development of Sensor Integrated Robotic Hand

Sahu, Om Prakash (2017) Design and Development of Sensor Integrated Robotic Hand. PhD thesis.

[img]PDF (Fulltext is restricted upto 16/08/2019)
Restricted to Repository staff only

7Mb

Abstract

Most of the automated systems using robots as agents do use few sensors according to the need. However, there are situations where the tasks carried out by the end-effector, or for that matter by the robot hand needs multiple sensors. The hand, to make the best use of these sensors, and behave autonomously, requires a set of appropriate types of sensors which could be integrated in proper manners.
The present research work aims at developing a sensor integrated robot hand that can collect information related to the assigned tasks, assimilate there correctly and then do task action as appropriate. The process of development involves selection of sensors of right types and of right specification, locating then at proper places in the hand, checking their functionality individually and calibrating them for the envisaged process. Since the sensors need to be integrated so that they perform in the desired manner collectively, an integration platform is created using NI PXIe-1082.
A set of algorithm is developed for achieving the integrated model. The entire process is first modelled and simulated off line for possible modification in order to ensure that all the sensors do contribute towards the autonomy of the hand for desired activity.
This work also involves design of a two-fingered gripper. The design is made in such a way that it is capable of carrying out the desired tasks and can accommodate all the sensors within its fold. The developed sensor integrated hand has been put to work and its performance test has been carried out. This hand can be very useful for part assembly work in industries for any shape of part with a limit on the size of the part in mind.
The broad aim is to design, model simulate and develop an advanced robotic hand. Sensors for pick up contacts pressure, force, torque, position, surface profile shape using suitable sensing elements in a robot hand are to be introduced. The hand is a complex structure with large number of degrees of freedom and has multiple sensing capabilities apart from the associated sensing assistance from other organs. The present work is envisaged to add multiple sensors to a two-fingered robotic hand having motion capabilities and constraints similar to the human hand. There has been a good amount of research and development in this field during the last two decades a lot remains to be explored and achieved.
The objective of the proposed work is to design, simulate and develop a sensor integrated robotic hand. Its potential applications can be proposed for industrial environments and in healthcare field. The industrial applications include electronic assembly tasks, lighter inspection tasks, etc. Application in healthcare could be in the areas of rehabilitation and assistive techniques.
The work also aims to establish the requirement of the robotic hand for the target application areas, to identify the suitable kinds and model of sensors that can be integrated on hand control system. Functioning of motors in the robotic hand and integration of appropriate sensors for the desired motion is explained for the control of the various elements of the hand. Additional sensors, capable of collecting external information and information about the object for manipulation is explored.
Processes are designed using various software and hardware tools such as mathematical computation MATLAB, OpenCV library and LabVIEW 2013 DAQ system as applicable, validated theoretically and finally implemented to develop an intelligent robotic hand. The multiple smart sensors are installed on a standard six degree-of-freedom industrial robot KAWASAKI RS06L articulated manipulator, with the two-finger pneumatic SHUNK robotic hand or designed prototype and robot control programs are integrated in such a manner that allows easy application of grasping in an industrial pick-and-place operation where the characteristics of the object can vary or are unknown. The effectiveness of the actual recommended structure is usually proven simply by experiments using calibration involving sensors and manipulator. The dissertation concludes with a summary of the contribution and the scope of further work.

Item Type:Thesis (PhD)
Uncontrolled Keywords:Sensors integration; intelligent robotics hand; parts identification; grasping points
Subjects:Engineering and Technology > Mechanical Engineering > Robotics
Divisions: Engineering and Technology > Department of Mechanical Engineering
ID Code:8675
Deposited By:Mr. Kshirod Das
Deposited On:17 Aug 2017 10:33
Last Modified:17 Aug 2017 10:33
Supervisor(s):Biswal, Bibhuti Bhushan

Repository Staff Only: item control page