In late 2016, my phone started slowly reaching "brick" status. The power button didn't work, the home button got stuck, and the battery started dying fast. My phone would heat up so much I had to take out of my pocket to cool down, the battery lasted only four hours. Unfortunately for me, school lasts seven, plus travel time of about half an hour each way. So I decided to come up with a solution.

My robotics mentor happened to have a couple broken school laptops on hand, so I asked if I could have one. Upon being granted ownership of a laptop, I took out the battery and very carefully opened the compartment. I then desoldered all the batteries and put them into packs of three cells wired in parallel. Afterward, I soldered them to a TP4056 battery charger and charged them fully to their nominal voltage of 4.2 volts.

Once I had the packs charged up, I hooked them up to a small LED and calculated they would take around two days to discharge the batteries fully. After two days, the batteries were fully discharged as expected.

Finally, I soldered the output of the battery pack to a five-volt regulator and attached that to a USB port. As a safety feature, I also attached a switch to stop charging in case of emergency. I also added LEDs for low battery, as well as when the pack is charging and fully charged.

During my freshman year of high school, I brought my Gameboy Advanced SP to my finals to play after I finished and waited to be dismissed. After finishing the two games I owned for the system I figured I would want to buy more games for it. Unfortunately, most of the classic retro games such as MegaMan 2 or Contra were not available on the Gameboy, so I decided to make my own Gameboy.

I figured that recreating a retro console using smaller components would have been a viable method but with today's upgraded technology those consoles are one of the worst ways to play a game. It's much more efficient and powerful to utilize a computer to run the games as a program, so I decided to buy a Raspberry Pi Zero and load it with the RetroPie operating system. To make it portable, I also bought some lithium batteries and a charging circuit. After directly soldering the battery output to the micro USB input on the pi, it booted up just fine.

The next issue was the screen, I wanted to find something cheap. The cheapest option (for about ten dollars on Amazon) was a rear view screen for your car. However this screen came with a little bit of a problem: it ran off the 12 volts DC from your car, not the standard five-volt output the batteries provided after regulation. I could hook up another regulator to the system, although that would be bulky. Fortunately, I noticed on the board there was already a voltage regulator IC. Even better, it's output voltage was five volts, so I detached the IC from the board, and soldered my battery wires to the IC output to replace the power. The only thing left to do for the screen was to solder the composite wire of the screen to the AV out of the Raspberry Pi and set the video mode to composite.

Afterward, I hooked up a board set up for push buttons to the GPIO pins of the Raspberry Pi. Thankfully RetroPie had an option to add GPIO pins as controller inputs, so I didn't have to code that myself. I followed the diagram found on their wiki on their Github page https://github.com/RetroPie/RetroPie-Setup/wiki/GPIO-Modules. After that, the only thing left in the wiring was to add a power switch, so I added a switch in between the battery output and the Raspberry Pi's power input.

Then came the hard part, making a case. I could've used an old Gameboy case, although my batteries would not have fit in the case, so I decided to make a 3D model for a case using Autodesk Inventor. I made a mount for the screen to hold it in place with hot glue, I added holes on the side for the charging LEDs and the switch. I also added holes for all the buttons and made the buttons separately in different files. With the help of my engineering teacher, I verified the measurements and printed the design.

For the most part, the project worked out really well. The battery life isn't anywhere near that of a normal Gameboy, but it lasts about five hours. The screen quality isn't awful, and the case still holds up. The only issue with the system that I still experience is the buttons, getting the buttons to work correctly is no easy task. Sometimes the buttons get stuck, and the left and right directional buttons won't register unless you press the upper portion of but pad.

This is a project I did with a few friends for a hackathon. We went to K-State and had thirty-six hours to program a project or "hack". Our idea was to make a bot that can write music. I started working on the backend while someone else started on the front end. We ended up coming with two different ideas for algorithms.

The first algorithm, which was written by myself: is a procedural generation of music. After selection of a key, and a bpm, and some other things, it uses a technique called cellular automata. Although usually used in a two, or sometimes three-dimensional sense, I used it in with a single dimension to create the rhythm of the notes that were to be played. Next, I used a probability curve to determine how many steps away the next note would be from the current one, as well as which direction.

The second algorithm was done in most part by someone else, but I'll still describe it. It's more of a "guided generation". It takes the Markov chain approach and samples notes from a different song, and uses those to determine the "best" placement for the next note(s). We weren't able to get this to work for every song, because we only had a limited amount of time to complete the project. So we manually inputted the notes from certain songs that you can choose to use as templates.

We had lots of fun making this project, and even more fun in presenting it to other teams as well as judges. At the award ceremony, we were awarded fifth place, which for our first hackathon we were more than pleased with. But after meeting the teams in ranks one through four, we realized we were the highest ranking team of high schoolers.

For the Java class I took freshman year we had an end of the year project, our task was to make a tic tac toe game. We were assigned into groups of three, but unfortunately, one member of my group graduated on day one of the project; and the other one was in the process of dropping the class, so I was on my own.

We got points for design, functionality, and extra points for making an AI. So using Java's "swing" GUI I made a basic tic tac toe game. I also added an AI, I wanted to make a good AI so I looked into some algorithms for doing this, most of them involved either really complicated math or machine learning.

However, I stumbled across one method that would work: the mini-max algorithm. So I wrote a recursive function that checked for a player or AI victory, and if there was no victory, it would call itself with a new move. If there was a player victory, it would subtract 100 points to a total move score, and if there was an AI victory it would add 100 instead. What that essentially did was calculate every possible move set left, and determines the move that results in the most likely victory. This method was inefficient, but the number of moves in a tic tac toe game is really low, if you tried to do this for a game like chess, it would take forever to do each move.

For a project for English class, I learned how to use Markov chains and applied to Edgar Allan Poe's The Raven.

In my junior year of high school, my AP physics would sometimes say he wished he had more simulations for demonstration purposes. So I made the effort of asking him what units he would want them for. This is the first one I made; I used the Unity game engine to make a simulation for illustrating the effect elastic and inelastic collisions.

This was the project I completed with others at the second hackathon we attended. After hearing about Youtube having to constantly take down videos because people who don't want to be in them threaten to sue; we figured it would be useful to have software that automatically does it for you.

This program uses python with opencv to process the image. Firstly it uses a pre-trained model in combination with some haar-cascades to detect any human face. After that, it tried to match it to any known face in the database.

To do so, I created a system that would recognize certain facial features on a person and add them to a database. When a face is detected in the program, it matches all facial features it finds to those stored in the database. If more match to a certain person than any others, it identifies them and doesn't blur them out.

The project worked really well, our only (and major) issue we had was optimizing the program. The video you see above is only about ten seconds long and took about five minutes to process. In addition, the UI isn't very well thought out. Currently, we're working on making it into a desktop application, as well as optimizing the program to make use of multithreading and Cuda cores in graphics cards.

The prevalence of mental illness and suicidal tendencies in our society creates a need for more help than ever. The National Suicide Hotline is a well-known resource for suicide intervention, but sometimes there is a long wait list for people who really need to talk to someone.

See more at https://devpost.com/software/project-pyper