Farmaid Bot

My proudest achievement so far. The award winning Farmaid bot was designed to help out arond small greenhouses at a very affordable price. Project Uses Python, Tensorflow, OpenCV, Robot Operating System, Arduino C, and C++ 11. The Project was self-funded and made in conjention with an independent team of engineers

Learn more

Health Promotion and Human Performance

I came from a place of no interest in sports to developing a machine learning projects used for monitering collisions in ice hockey and publishing three research papers on it. Technology used includes Python, OpenCV, Tensorflow, Numpy, Pandas, PyTorch, Microsoft Excel, NodeJS, and Bluez This was part of my position as Graduate Research Assistant at Eastern Michigan University which I recieved after my first semester of my master's degree

Learn more

College of Technology

Part of my Practicum course at Eastern Michigan University. I took this position in place of a higher paying one because it offered more challange and a chance to work more with robotics. Projects included the Donkey Car based robot and the ML model for movement tracking application for iOS all done. Technology used included Python, C++, OpenCV, CMake, Tensorflow, ONNX, Caffe, and CoreML

Learn more

Projects

Here are some detailed descriptions for my recent projects

Farmaid Bot 1.0

Built in response to the ARM Autonomous Robot Challenge, in summer 2018, I first wrote a small pitch which won a Raspberry Pi based Donkey Car robot. Then, I reached out to the Detroit Autonomous Vehicle Group to see if anyone with more electrical experience wanted to join me and was contacted by people with various specializations and over the course of July 2018 to September 2018, we set about creating our project. I was responsible for most of the programming and some of the wiring and hardware. At the end the project won the award for Best Use of AI. A smaller version of it was built at Hack Harvard in October 2018 and won Most Fundable Hack award.

Learn more Back to top

Farmaid Bot 2.0

We gained some momentum from the last victory and I used it to pitch an update to the project at Optimize Eastern, a student group that helps fund student enterperneurial projects. Here I won some funding and set about creating Farmaid Bot 2.0. Based on the success of the previous project, the Detroit Autonomous Vehicle Group was able to do more marketing and attracted more volunteer researchers including an expert in Robot Operating System from Ford and an electrical engineer from Chrysler, while I was able to focus more on the computer vision tasks and move from classification to an object detection method.

Learn more Back to top

Impact Detection

I was given the problem of detecting collissions in ice hockey players and some tablular data along with videos of the game the data was extracted from, the restriction was that I was only allowed to use 8 columns from the 100+ column dataset as the other columns were collected at 100hz for each player and increased the file size too much, lowering the size was first priority. For the size reduction, I created a method using Autoencoders, Support Vector Machines and Multi-Layer Preceptron and was successfully able to lower the size while still getting good predictions. I also proposed a method to detect impacts via video data and crated a dataset to corelate the videos to the tabular data and used an object detection based method to do detections and applied additional filters to remove false positives. The tasks resulted in two research papers being published in the American College Of Sports Medicine(ACSM) Journal.

Learn more Back to top

Donkey Car

Since the chassis of the donkey car I won from Farmaid was not suitable for a greenhouse environment, I donated it to a professor at a different department who ran a meetup group to showacae student projects. I worked with him during my practicum course to train a model that could drive autonmously around a track and also on training a model via a simulated environment made in the Unity game engine. The project was displayed at the North American International Auto Show in Detroit.

Learn more Back to top