Past Projects (2016-2020)

Motorcycle (2020)

I purchased a motorcycle with a leaking fork and extremely old tires. I was able to diagnose the fork issue down to a leaking fork seal. After this, I completely disassembled both front forks, replaced their oil, cleaned them, and replaced the fork seal with a replacement kit and reassemble the bike. This has fixed the issue entirely. I also swapped out the front and rear tires of the bike myself as they were cracking and unsafe. While this isn’t particularly an “engineering” project as it is solely mechanics, I still feel it demonstrates one of my hobbies as well as showed me how to use torque wrenches, how to service engineering work, and allowed me to appreciate the intricacies of such a complex machine.

FRC Year 2 (2019)

This robot is from the 2020 FRC game where I was the mechanical team captain. We tried a much more robust implementation of sensors into essentially every subsystem possible for closed-loop control. 

This robot was the last robot that I ever built, it was in my final year as mechanical captain of my high school robotics team. It was designed to essentially pick up 5 balls approximately the size of dodge balls, index them, then shoot them. As the mechanical captain I was in charge of essentially the entire design and assembly of the robot and subsystem integration. Specifically, I completely designed the shooting mechanism which uses two flywheels with sensors to measure wheel RPM and shoot a ball out. This was coupled with an off the shelf vision system to approximate distance from the goal and thus ramp up the RPM based on a distance to RPM curve that we experimentally found out. Along with this, I designed the entire ball indexing system, the two vertical and one horizontal runners with essentially belts that aren’t pictured in the CAD. 

This is a closer up image of that system. The reason why this was necessary is because dodge balls, when they interact with each other and are squished, they create a lot of friction and essentially become sticky. Thus, before launching you need to serialize the balls so it doesn’t jam your shooting mechanism. This design basically did that by having the left and right vertical runners run in reverse essentially trying to push balls out while the bottom runner pushes balls in. This prevents any balls from getting stuck together and serializes them. 

Finally, on this robot I also led the design of the drivetrain. The drivetrain design is called a west coast drive, designed originally by west coast robotics teams, with an out of tube chain format. This design originated from a drivetrain that our team made a year earlier (my first CAD design ever and the drivetrain found on the 2019 robot) which had a chain-in-tube design. We opted for a chain out of tube design here because it is much easier to maintain and service as well as allowing larger chain sizes for larger wheels. This was my first lead on a project with about 7 people working under me in the summer design process. 

During the actual build of the robot I led a team of around 20 people which was quite hectic at times. We were a team with very little high level members and a lot of new freshmen members so I had to work around minimal experience during the design process and build as well as manage that many people. 

Overall, I really liked this robot design. There were some kinks that needed to get ironed out over time and it’s one competition that it competed in it did not perform well due to those problems. I never got a chance to fix those issues due to COVID but I like to think that this robot would’ve been successful at its tasks. The main cause of these issues was manufacturing problems. Between the 2019 year and 2020 year the team lost some senior members who worked at machine shops and were able to machine parts for the robot. This led to completely in house machining which for some parts led to poor meshing and interaction and overall excessive power consumption from motors because systems didn’t align perfectly. These are issues that would’ve likely worked themselves out with time but unfortunately time was not on the table due to COVID. 

I think more than anything this year of robotics taught me so much about leadership and being a good leader. This was the largest group I ever had to lead toward a common goal and that was the most difficult part of the entire robotics season.

Cinematic Quadcopter V2 (2019)

FRC Year 1 (2018)

This was my first year on a robotics team, my Junior year of high school. The team was where I learned how to CAD as well as applications of engineering. So on this robot I designed the drivetrain (this was designed in the off season and where I actually learned how to CAD originally), the arm linkage, and an end effector which isn’t shown in this CAD model. 

The arm linkage was a virtual 4 bar design, this design was chosen because the objective of this robot was to pick up small yoga balls and lexan plates and then place them at varying heights. This required our end effectors to always be perpendicular to the floor, thus the necessity for a 4 bar mechanism. The main issue with this mechanism was the amount of slop in the system because, while hard to see in the CAD, the 4 bar is mounted onto a lift system which gave a very small footprint for the entire 4-bar linkage to take up. This forced the entire linkage as well as motor drivers to fit in a small area limiting the way that I could setup the connection between the motors and arm. I opted to implement a chain drive system connecting the 4-bar linkage to the driver motor but this was designed late in the process and was an overall poor design. There was no implementation of a robust tensioning system which caused a lot of slack in the arm system, especially considering the arm was rather heavy and holding mass at the end of the arm, which made the arm very difficult to consistently control. This was really the main problem with this robot and if I were to do the entire thing again I would have opted for a 4-bar linkage powered by a piston for robust control for two positions considering we only really needed the 4-bar to access the top heights of the challenge. For a visualization of the challenge for this year just google “First Robotics Competition 2019 Challenge”. 

This is an image of the end effector mentioned earlier. This end effector was designed to essentially grab onto a lexan plate with a hole in the center using those 3 central petals. Once the lexan plate needed to be released, those petals would revert back to their original position and the two side pneumatic pistons would fire. This would ensure that the lexan plate was securely attached and also allowed for some degree of error in alignment. The end effector didn’t have to be perfectly aligned with the interface for the system to work, as the pneumatic pushers would account for around a 20 degree alignment error either way. 

Overall, this robot was not great. There were a couple weak points that really caused the downfall of this robot. Notably the 4-bar linkage as well as the lack of integration of sensors into the robot for closed-loop control. 

Quadcopter V1 (2016)

This was the first big electrical project I ever undertook. I began this my freshmen year of high school with the objective of building a cheap and simple drone that would teach me the basics of soldering, electrical assembly, and working with RC aircraft. As I was attempting to make this as cheap as possible, the motors are not quadcopter motors but rather fixed wing RC plane motors thus two of the propellers could technically fall of mid flight (two motors spin in opposite directions to counteract drifting due to conservation of momentum) because they are not threaded in reverse (this has yet to happen). This project also initially started as a failure, its first flew flights induced huge growing oscillations that flipped the quadcopter crashing it into concrete. However, a few years later I realized the issue was due to poor mounting of the flight controller leading to excessive vibrations which caused the oscillations. This was because the flight controller was attempting to react to the vibrations due to poor mounting and oscillating the craft. I fixed this issue and now the quadcopter flies flawlessly. I learned so much through this project and it drove me to pursue further electrical and mechanical projects in the future.