1. Conclusion

Our project demonstration went really well as our mechanism never failed to sort a ball in the correct location. The main goal was achieved of minimizing acceleration with an acceptable byproduct of a slight jitter to the motion of the vertical movement. The jitter was minimized as much as possible as allowed by the resolution of the servo available to us.

2. Future Works

Future improvements mainly consist of adding feedback controls as we had no sensors on the main mechanism. An ultrasonic sensor would help our horizontal movement to be more precise and accurate. A touch or bump sensor would also allow for us to confirm if the object has actually been picked up or dropped. With no feedback, we had to rely solely on time. Different vertical and horizontal movement mechanisms could also be explored to determine which one would be more precise and scalable to industry applications. A feeder mechanism to remove the human input of balls would also automate our system even further and make it a more complete package. While an integrated feeder would remove some human input, industry applications would possibly not require a feeder as they might already have their own dedicated systems.

Other noteworthy aspects of our final design are the inconsistency that a screw drive brought. This was derived from mounting the servo to the screw itself as any bit of misalignment would cause the screw to shift, and thus our mechanism to be at the wrong starting position.

3. Lessons learned

• Actuator resolution can limit motion smoothness even when acceleration is minimized.

• Feedback control is critical for accuracy and repeatability; time-based control is less robust.

• Mechanical alignment and tolerances strongly affect system reliability.

• Early integration and testing reduce debugging time near project deadlines.

4. Tips for Future Students

A tip for future students is to make use of TIW resources as early as possible as they have set closing times and close shop on the last day of classes. Lastly, finish the final build as soon as possible as we spent the most amount of time debugging our code.

5. Acknowledgement

We want to thank TIW for their assistance and machines. We want to thank Dr. Christina Petlowany Cade Wetherill, and Min-Geun Park for guidance.