This journey brings us to today and the current state of this research group. With Project Zeta, MAGIST, and RaSDAN, there have been huge milestones accomplished any many more yet to come. If you want to contribute to this journey, please check out my GitHub and our projects and feel free to get in contact with me.

Smaller, faster, and more security features, this devices was the final design for the CarLink module and was tested on multiple vehicles!

This is the first DeepShift Labs project with a online 3D design, custom motor housing and custom software. This was meant to be a extensible robotic platform but had many dynamics issues that would later be resolved.

On a journey to improve the performance and reliablity of autonomous vehicles, significant research was done for the CarLink project. Doing what the US Department of Transportation could not, I developed a Vehicle-to-Vehicle communication system that is fast enough, accurate, and reliable to allow autonomous vehicles to know the exact location and intent of regular vehicles and alert regular vehicles of dangerous drivers.

Lessons learned from the first version were applied with stronger motors and a custom protoboard that connects the microcontroller on the robot itself. This robot performed much better with the major downflaw of the weaker motors. This would be the stepping stone to Project Zeta.

To see the max potential of wheeled robots, I decided to research holonomic motion. Swerve drive was too complicated on such a small form factor and mechanum was too difficult to effectively 3D print so I decided to create a 4 wheel omni-wheel X-drive!

Tired of using wheels, I started my very first experimentation with quadrupeds. This used an Arduino Nano and 12 9g servos and was too weak to carry anything so the control circuit was tethered from the outside!

Using new learned techniques and a broken controller from an old toy, it was repurposed and soldered into an Arduino and nRF radio based all-purpose controller. This provided a must more solid and robust method to control my various robots.

This was more experimentation with human-machine interfaces as this contained a stereo microphone and headphone speakers on each ear. With a Raspberry Pi and a Pi Camera, I could make internet searches, cast videos to our TV, and detect objects and people with the camera all in a custom integrated Google Assistant!

After many failed flights and broken propellers, this is the flight testing that was done on the drone. After several more revisions from here will I arrive to the final design. At this point, I have move away from using breadboards to a custom circuit board I soldered on some protoboard.

With bluetooth too slow and WiFi on mobile phones to rigid, I decided to make my own custom controller. This controller is based on the ESP32 and uses a custom protocol to communicate with the robot. This project used many input methods connected via an Arduino Nano and a nRF24L01+ radio module. Future revisions would soon include BLE, WiFi, and a 2.4in TFT Touch display.

Four wheels are boring so here is another side excursion with PID controlling a 2-wheeled robot. Using bluetooth and a MPU6050, this robot can be controlled with a phone and can balance itself. This project was a great way to learn about PID control and the importance of tuning.

A quick side excursion became a very interesting project. This project proved the extensibility of old or unapplicable electronics as this project uses steppers and carraiges from DVD readers and the steppers are driven by H-Bridge DC motor controllers! It further utilized Marlin GCODE parser and Inkscape bitmap maker.

A project that would go down as one of the most difficult projects I have ever done! With completely custom hardware and a 10lbs. payload capacity, this behemoth has crashed far more times than it has flown and has proved to have many milestones in electrical and control theory understanding. The knowledge gained from this project about PID and control theory as well as electrical stability in high-vibration environments would prove useful in Project Zeta.

This was the first robot packed with sensors and equipment. It had a pivoting ultrasonic range finder, a camera, and a speaker with a Raspberry Pi 3B+ running a custom version of Google Asisstant. It can run fully autonomously and avoid obstacles, with a gesture sensing glove, or with a custom designed controller. Additionally, the camera could directly FPV stream to a VR headset.

When I was in 4th grade, I recieved my first Arduino kit. This would be the necessary push to kickstart my interest in robotics and AI. I remember spending hours trying to turn on an LED and spin a motor. Through the frustration, however, arose the need to persist through challenges and would form the foundation for project today.