...
Our next idea was to develop a crossed four-bar linkage configuration that was found to mimic the curved motion of a human finger. We based our model off of a finger mechanism patent described in the kinematics section and developed a curved pattern that is similar to a curved motion of a finger.
Figure 2: Crossed Four-Bar Mechanism
...
When starting our prototype, we began creating the mechanism with cardboard to map out the movement with physical pieces. We measured the the link lengths to approximately the desired lengths that were defined in Motion Gen. The goal for this prototype is to observe and understand the motion of the mechanism using physical components. However, cardboard is very flimsy and susceptible to breaking down after rep. 
Figure 3: Cardboard prototype of the open 4-bar mechanism
...
Prototype 1 uses a bracket with a screw to lift the tab. This was the first material we thought of trying as it is thin and sturdy on the links.
Figures Figure 7: Initial Design design of the End Effectorend effector
End Effector Prototype 2:
After testing prototype 1 for the end effector, we found that the bracket is not strong enough to withstand the force needed at the tab. We saw that the bracket started to warp and much of the force from the links is being transferred into the deflection of the bracket instead of onto the tab. We switched out our bracket system and machine machined a steel piece that is both thin enough to get under the tab and strong enough to pull the tab.
Figure 8: Laser Cut Linkages With Custom End Effector cut linkages with custom end effector
End Effector Prototype 3:
The last prototype was to fix the new end effector to the acrylic links. Below is the final iteration of the end effector piece.
Figure 9: Final End Effector end effector with acrylic linkages
Prototype 4: Final Iteration
With the previous iteration, we found that the torque output from motor does not supply enough force to open the can. Additionally, we currently don't have linkage system at the correct height to reach the top of the can. We addressed the torque issue by created a 49:1 gear systems from the motor shaft to the rotating link to overcome the lack of torque output. Because of the size of the gears, we elevated the system using two acrylic sheets, which addressed the height problem we were also having.
Figure 10: Final Iteration