Customer Handoff & Final Project Documentation
Table of Contents
Team Vision for Final Demo and HandoffTo Complete:
- Replace servo motor broken during previous demo
- Disassemble extension mechanism, loosen hardware for smoother sliding, re-apply loctite
- Finish engineering requirement testing
- Finalize poster & technical paper
- Present robot at ImagineRIT
- Wrap up any unfinished documentation
- Clean up EDGE from outdated information
- All of above, however the replacement motor did not provide enough torque to drive extension mechanism reliably
Test Results Summary
Link to video showing robot successfully operating underwater:
Deployable Depth Test
Robot was deployable to at least 2.4 m. Limited by length of control panel tether.
Life Cycle Test
Gripper successfully actuated 120 cycles during test without issue. The maximum number of cycles were not reached, as the team did not want to take the gripper to failure. However, there were no signs of degradation during testing.
Successful Capture Rate
During testing, the test object was successfully gripped 100 percent of trials.
Performance vs. Requirements
Risk and Problem Tracking
Final Project Documentation
CAD/SchematicsSolidworks CAD Files and Drawings (ZIP Archive)
Final BOMFinal BOM
SoftwareThe final version of the software can be found here. It should be noted that the final software does not include any pressure sensor code as the pressure sensor is no longer functioning.
Recommendations for future work
This project successfully showed that soft robotics can be taken underwater and used to grasp objects. The grasping device operated reliably throughout testing and demonstration. With further development and exploration of gripper geometry, this technology could be attached to an underwater exploration vessel and successfully used to collect fragile specimens underwater or other similar use cases.
The main point where this project did not operate as intended was with the extension mechanism, which did not actuate as smoothly as intended. The servo motors used did not provide enough torque to overcome the friction from the scissor mechanism. To improve this design, use of a double-acting hydraulic cylinder is recommended, as it can be driven from the same pump as the soft bending actuators. Since this is a very low load application, an industrial hydraulic cylinder is not recommended due to cost, but rather a low-cost design that would be easy to fabricate. Another option is to mold the extension mechanism out of a soft material as well, with a bellows style design, but this could present issues with the capacity that the robot could lift.
Also, the robot was limited in the size and weight of objects it could grasp due to the geometry of the manifold and actuators. Any design will have similar limitations, but to make the design more adaptable without having to fabricate multiple manifolds, it is recommended to mold the entire end effector out of a single mold, with a single hydraulic inlet. Similar designs have been pursued by the Whitesides research group at Harvard. This would not only make fabricating a variety of gripper geometries simple, but would reduce cost and prevent potential issues with connection between the manifold and actuators.
Future improvements to this technology would allow the robot to grasp objects with a wider range of geometries and weights, as well as sense when an object is grasped. Integrating sensing into the gripper device by molding sensors into the base layer of the soft bending actuators would allow object grasping feedback. Also, attaching the gripping device to a multiple DOF arm would increase its usefulness in underwater applications.