Final Test Rig
Skin Testing
This was a project in collaboration with the startup robotics company, Sensible Robotics. I worked as an Engineering Consultant on a project where we were tasked to research, develop, and test silicone skin designs optimal for humanoid robotics. Our three main objectives were:
1. Create an efficient test rig for skin design iterations
2. Design and mold the skin iterations
3. Use Test Rig to test Skin Patterns through 2 procedures to measure skin pattern efficacy in common situations.
I was in charge of creating the mechanical test rig which would enable our skin design team to test their different patterns and materials while keeping these constraints in mind:
Challenge 1: Develop a mechanism capable of testing the gripping action of objects in a variety of orientations.
Challenge 2: Ensure equal and consistent force across fingers when applying the “pinching” motion.
Challenge 3: Ensure replicability of all settings of the test rig.
Our Solution
Initial brainstorming led to us to create this underactuated gripping mechanism that allowed for us to reiteratively test the skins while only having only one test variable, the angle of attack of the fingers. This would mean that the only thing we would be changing between tests would be the angle at which the fingers were coming together and gripping our test objects.
Initial Drawings
Final Design
Annotated Test Rig
To ensure an equal application of force across two fingers we developed an underactuated gripper mechanism powered by a single spring to provide a calculable force to the objects gripped within the function.
To create a wide range of test configurations that were easily replicable, Hirth Joints were employed to allow for rapid reconfiguration to a series of set points.
Claw Mechanism
Claw In Action
Final Test Rig
Our claw mechanism was closed at a natural state and could be manually opened to grip onto different test objects. The fingers were required to be parallel throughout the actuation of the test rig.
Initial Finger Mechanism
Final Finger Mechanism
Finger Manufacturing
Our finger adjustment design was originally going to be motorized, but we felt this was inaccurate when acted upon by a force probe. We then decided to attach the fingers given to us by Sensible Robotics via screws to the Hirth joint plates, but we later incorporated the finger directly into the gears and resin-printed it entirely. This allowed for us to mass-produce these fingers and apply the skin designs (red silicone) for testing.
Sliding Casing
Removable Casing
We were also tasked with creating a mechanism that allowed for variable forces applied to the test objects. To do this, we used a variable spring force mechanism that allowed us to calculate the force being applied by the testing rig onto our test objects. Since the testing rig is closed at a natural state via the spring, we could use a mechanism to lengthen this spring and increase the force applied to the object using Hook's Law. We went through multiple iterations involving sliding casings, removable casings, and finally, a wing-nut actuated mechanism.
Wing-Nut Mechanism CAD
Wing-Nut Mechanism
Our final design utilized a wing-nut to lengthen the spring and increase the force. We included markings of 0.2 in and 2mm on the sides of the casings to calculate the spring's displacement and calculate the force applied. The spring was attached to a wooden dowel which was attached to our actuator.
360 Movement
Final Mount Assembly
Rig Movement
We used a Bon Tool mechanism to allow for 360 rotation of the entire testing rig. This would allow us to test different angles in which the fingers were gripping onto items and where the force of gravity would be applied to the objects from a different angle. We also created a plywood base and a 3D-printed insert for more stability.
Skin Molding
Force Probe Testing
Our other team members created multiple skin designs inspired by common sports grips to biomimicry. We resin-printed molds and used silicone molding to produce the skins which we then glued onto our fingers. We tested 4 different skin patterns and their performance in gripping onto metal, plastic, rubber, and wooden test objects. To test the grip strength of each skin, we used a force probe to push out the objects from the fingers and recorded the highest force measurement.
Results
Force Readings by Material and Skin Type for Cube
Force Readings by Material and Skin Type for Sphere
Our tests resulted in a clear winner of a humanoid fingerprint pattern that allowed for a heightened grip on metal, plastic, rubber, and wooden objects.
