19.6 Conclusions & Future Work
Overall, our project was a success. We successfully generated enough impulse to launch our rabbit into the air using a combination of a cam, a four-bar linkage, and a torsion spring system. While the resulting motion more closely resembled a bucking rodeo bull than a natural rabbit hop, we were still pleased with the device’s ability to achieve lift-off.
Lessons Learned
Technical Side
Asynchronous Servos
We used two separate motors in an open-loop control setup, without any feedback mechanism. As a result, the robot struggled to jump with both legs simultaneously. We spent a considerable amount of time manually tuning the physical assembly structure in an attempt to synchronize the servos. This synchronization issue could have been more efficiently resolved through the use of a mechanical transmission system, but at the cost of a more complicated transmission design and motor power management. Alternatively, implementing position sensors, such as Hall effect sensors, would have enabled a closed-loop control system to ensure better coordination.
Design Priorities (fear of not jumping)
While we initially treated the total weight of the robot as a primary constraint — to the point where we had to consider detaching the power source and make the system dependent — it turned out that minimizing weight was not as challenging as we anticipated. Instead, we could optimize jumping performance — including impulse and height — through design improvements in the four-bar linkage and cam-follower systems.
Buy Smart, Not Cheap
One of the biggest takeaways from this project was learning not to prioritize low cost at the expense of performance and reliability. In an effort to cut costs, we purchased from Amazon and didn’t realize that those were solely left-handed torsional springs. It was too late when we realized at assembly, after everything was fabricated. Although we managed to adjust the design on site, the asymmetric torque distribution from the identical spring orientation introduced a structural imbalance. This imbalance of one leg sitting higher than the other persisted when the robot was powered on and in motion. Moreover, the inexpensive torsional springs lacked specification sheets, forcing us to spend a significant amount of time experimentally determining their properties needed for accurate kinematic analysis. We also chose low-cost chips and motors without purchasing extra spares, both of which unexpectedly failed and had to be replaced in a hurry. These failures introduced delays and complications that could have been avoided by investing in more reliable components from the outset.
Project Management Side
The biggest lesson we learned from this project was the importance of starting early. Throughout the semester, we often became a little too comfortable in the weeks leading up to preliminary build assignments, which resulted in a few last-minute rushes to complete our builds and analyses. However, by the time the final project approached, we were much more diligent about making steady progress week by week. This made the overall process far less stressful, even though some last-minute work was still unavoidable.
Another key lesson we learned was the importance of clear communication between subgroups. Early on, we split into analysis and design teams to work more efficiently, but this sometimes led to misalignment—design decisions were made without input from the analysis side, and vice versa. These disconnects caused delays and rework. As the project progressed, we improved our coordination through more frequent check-ins, which ultimately led to a smoother, more cohesive process.
We also learned the value of actively seeking guidance from available resources, particularly Dr. Symmank, Connor, and Mila. In the early stages, we didn’t always come prepared to weekly check-ins, which limited the value we got from those meetings. As the project progressed, we made a point to show up with specific questions and clear updates, which led to more productive discussions and valuable feedback. Their input helped us avoid missteps, refine our design, and ultimately deliver a stronger final product.
Future Improvements
Purchase quality components:
Investing in quality components is essential. Major issues we encountered were due to limitations and inconsistencies in the low-cost components. By selecting reliable parts from the beginning, we could have avoided spending time troubleshooting problems that stemmed from the purchased products, not the design or software.
Improve the cam-follower system:
We can optimize the cam-follower profile — specifically, the shape of the cam — to reduce energy loss in order to minimize energy dissipation during the energy transfer from motor to cam and help amplify both motor output and the overall system response.
Further system analysis:
To achieve a more realistic and accurate system analysis, we should further our understanding of mechanical system modeling and integrate the analyses of the cam-follower mechanism and the four-bar linkage. Combining these two will provide a more comprehensive view of the dynamics, better matching the actual system behavior.
Smarter control
Incorporating remote actuation so that we can trigger the rabbit’s hop wirelessly. This would offer a finer degree of control of the rabbit as opposed to the cams just spinning continuously, causing the rabbit to hop repeatedly. Future iterations could also explore other, more advanced movements, such as backward hopping or even flips.
Clearer System Identification
Another thing we realized only towards the end was that we had been mapping our design on top of a four-legged animal the entire time, while in reality, our system resembles more of a hopping bird or dinosaur-ish reptile, since the front legs are passive. Thus, the weight distribution was inevitably front-favored despite tuning battery weight on the tail. We should’ve identified that earlier to better inform our design choices
Acknowledgments
We would like to thank Dr. Symmank, Connor, and Mila for excellent mentorship throughout the semester. Their guidance was instrumental in helping us navigate design challenges, refine our ideas, and bring our project to completion. Also, special thanks to the TIW staff for additional assistance during our prototyping process.
We would like to acknowledge ourselves for choosing a challenging topic—not only achieving our goal, but also gaining valuable learning experiences throughout the development process.