13.7 Conclusion and Future Goals
Conclusions & Future Work
Project Outcomes
The project successfully demonstrated that an inverted slider-crank combined with elastic energy storage can enable jumping behavior using modest actuators. The design met its primary objective of achieving a jump through mechanical amplification rather than raw motor power.
Major Limitations and Failures
High torque during loading: Stretching the rubber bands required significant torque, pushing the limits of the DC motor and pulley system.
Asymmetric loading: Small manufacturing imperfections caused uneven resistance between legs, leading to biased bending and uneven jumps.
Structural bending: String-based loading introduced off-axis forces that caused members to bend under high tension.
Clutch failure: The 8 mm interference-fit clutch stripped under high torque, eventually spinning freely on the shaft and limiting repeatability.
Lessons Learned
High-force mechanisms are extremely sensitive to alignment and symmetry
Interference fits with 3D Prints alone are insufficient for high-torque clutch interfaces
Elastic energy storage simplifies actuation but complicates structural design
Manufacturing tolerances matter significantly in near-singular mechanisms
Calculating torque and power required for actuators are essential to have a function robot
Future Work
Smaller Design Changes
Replace interference-fit clutch with keyed or splined shaft
Improve structural rigidity to prevent bending under load
Add symmetric energy storage paths to balance loading
Introduce sensors for preload measurement and closed-loop control
Optimize linkage geometry to reduce required preload torque
Big Design Changes
We noticed that our intermittent gear mechanism designed to disengage and engage the clutch (using a slider crank mechanism) was very unreliable in the final prototypes, and does not have the capacity for handling forces required for higher jumps. We would likely implement linear actuators (solenoid, linear screw actuator) as it can come in a compact high force package and not have the issue of creating a radial load on the shafts on which the motor moves
Additionally, in the future, we would look toward implementing forward, backward, and side to side motion to add another form of locomotion in conjunction with jumping. We can also control the jumps in the future through the use of an accelerometer (similar to Boston Dynamics Sand Flea Jumping Robot).
Tips for Future Teams
Design joints and shafts assuming worst-case torque
Prototype high-load interfaces early
Avoid string-based force transmission when alignment matters
Test elastic elements experimentally, not just analytically
Minimize friction within the system through use of bearings (since high loading causes internal forces
Acknowledgements
We thank Dr. Petlowany, Cade, and Min-Geun for technical guidance, support, and feedback throughout the project. We would also like to thank the staff at TIW and Scott Evans for helping with the manufacturing of this assembly.