Tree Climbing Quadruped Robot
The aim of this bachelor thesis is to Design and Develop a Tree Climbing Quadruped Robot. The 12-DOF (Degrees of Freedom) Quadrupeds developed have proven their dexterity in traversing a variety of terrains and such robots find their application for operations in a hazardous environment, disaster rescue operations & defence. But these bio-mimicking robots will only be able to go to places with direct physical access, like a road or unstructured terrains, and stairs, etc and in cases perform some jumps and flips. Enabling them to take the support of existing structures and make their way will increase their applications by manifold.
This project primarily targets the agricultural application of the robot in fruit harvesting. Many fruit-bearing trees are structurally weak to support a human. Traditional methods of harvesting fruits using sticks are cumbersome and quite time consuming as it is difficult to reach some fruits even with long sticks and it involves the risk of damaging the fruit during harvest. Using human labour for such process poses a threat to the very life of the human being. So, the objective has been set to develop a quadruped robot to climb over the tree. The robot will not only be able to climb the trunk, as in case of many wheeled coconut tree climbing robots but will also be able to transfer from trunk to branches and scale whole length and breadth of the tree. Out of the options available, 11-DOF robot including four 2-DOF legs and a 3-DOF spine has been designed and developed.
With 11 DOF this robot essentially mimics a human with only upper leg and upper arm, supporting himself on his elbows and knees. Trying to walk in such a situation, we can observe that in this case, the spine plays an important role to stabilize locomotion gaits by shifting CG inside the support polygon that allows one of the leg to be lifted and placed in the next position. Therefore, the active spine has been brought into action to stabilize the locomotion gaits. One of the videos shows the improvement in the stability when the spine motion has been incorporated in the locomotion gait.
The robot developed has an open-loop control system with scripted gait. Currently, the work is going on to integrate sensors and camera for better locomotion with environmental awareness. ROS is being integrated in the raspberry pi controlled robot for better management.
The project stood runner-up at the Samsung Innovation Award 2019, at IIT Indore and won the prize of INR 120,000/-. It has also been nominated for the best B. Tech. Project award at IIT Indore.
I am highly grateful to Dr. Devendra Deshmukh for supporting my endeavours even though out of his research area and providing me with much-needed guidance and support.
Thesis available at:
http://dspace.iiti.ac.in:8080/jspui/handle/123456789/2177
Coverage in Media:
1. Samsung Newsroom India
2. The Times of India
3. Asian Age
4. Outlook India
https://www.outlookindia.com/newsscroll/samsung-recognises-iitindore-students-for-innovation/1660227