Project Lead - November 2021 to May 2022 

My partner Fernando and I led a team of 4 to design, fabricate, and assemble a mini UAV and UGV to be piloted in a competition representing the University at Buffalo. 

Design, build, and pilot a small Unmanned Aerial Vehicle (UAV) to transport an Unmanned Ground Vehicle (UGV) through a series of obstacles to perform reconnaissance mission, and navigate the UAV back to the initial starting point.

The UGV must be capable of detaching from the UAV using an autonomous onboard detachment mechanism, and autonomously drive to a different location after delivery. The mission must be completed under 4 minutes.

Between Fernando and I, we were responsible for: 

• Material selection for the wings, body, and legs. 
• CAD modeling of the UAV and UGV frame, as well as the connection mechanism. 
• Battery and electronics testing for the UAV and UGV. 
• Sensor selection, testing and troubleshooting. 
• Solder, test, and troubleshoot an I/O embedded system for our UGV. 
• Battery and motor selection via electronic and aviation simulations.  
• 3D printing and assembly of UAV and UGV.
• Create and submit a bill of materials. 

We decided to utilize line tracking technology to drive our UGV to the target platform. The competition allowed teams to purchase pre-built/designed kits, but we decided to design the ground vehicle from scratch so that we can personalize it to the sensors we wanted to use. 

We decided our system would consist of the following components:

1. Two Infra-red sensors: used to enable line-tracking.
   
2. Arduino: microcontroller to gather data and perform set actions.
3. Motor Driver: Used to control DC motors.

To minimize weight, I decided that we could use an Arduino Nano and the motor driver DRV8833. They met our requirements and were the lightest options in terms of weight. 

Additionally, the team decided to design a servo-based connection mechanism between the drone and the ground vehicle that would be controlled by the Arduino Nano. The infra-red sensors also work as a distance detector, so they can be used to tell the servo when to detach the car from the drone. The circuit schematic is shown below as well as the self-made PCB used to assemble the circuit:

The Infra-Red (IR) sensor is composed of an IR transmitter and receiver. When the sensor is near a white surface, the IR signal bounces off the surface from the transmitter into the receiver and the sensor outputs a signal of 100% the amplitude of the input voltage (5v in this case). When the IR sensor is over a black surface, the IR signal gets absorbed and it never returns to the IR receiver. Therefore the sensor outputs 0% of the input voltage.

Our UGV had two sensors on each side. The Arduino nano reads the output signals of both receivers and it tells the motors in which way to turn, i.e. if the right IR sensor is on black, the right motors must turn counterclockwise and the left motors must turn clockwise for the UGV to make a right and correct itself. 

Our code can be found on Github here 

The logic of the code is shown in the image below:

The Unmanned ground vehicle (UGV) was the first step for our team. Building it first would allow the team to get a rough idea of what is the total weight required for the drone motors.

The first part was to design what the connection mechanism between the drone and the UGV would look like. We decided to attach it to the top of the UGV. A visualization of how the connection mechanism works is shown below. The red object is meant to represent the bottom surface of the drone while the gray object would be on top of the UGV. This connection mechanism was modeled by Fernando. 

This prototype consisted of a breadboard and many many wires. This models main function was to test our electronics and power for proper functionality. 

This model was not feasible due to the vibration within the wires while the UGV moved forward. We moved forward by creating our own PCB. 

For this prototype, we utilized a PCB with soldered connections as a replacement for the breadboard, and most of the wires in the previous model. This helped us shed considerable weight, and made our design more resistant to vibration.

This model was not feasible for our final design because it did not contain the separation mechanism, and was too wide in nature to be carried by the drone.

For our third evolution, we fabricated and 3D printed a slimmer multi-level frame for the UGV. This design allowed for all of our components to be housed securely within the middle of the UGV. Also, this design allowed for the connection mechanism to be built into the top of the UGV.

This model was capable, but we ultimately decided on one more evolution in an attempt to refine our material selection and minimize final UGV weight.

Our final prototype had a modified frame, as well as utilized a different printing thread and material. On the bottom of the inside surface, we created indents for the battery packs to lay in, as well as modified the size of the connection mechanism. These changes decreased weight, and increased the stability of our components while the UGV is being carried, and while it drives.

We were satisfied with this UGV design and performance, and were ready to move onto UAV design.

Below is a video of our UGV executing the mission as intended. 

To begin the design of our drone, we needed to identify weight requirements and constraints. We created a table containing the weight of all our known components, and over-estimates of components we did not have weights for. 

With the inclusion of a safety factor of 1.2, we estimated that our UAV and UGV would weigh about 1800 grams, and that our motors should be able to handle this weight.

Once the weight was determined, we needed to determine the thrust that would be required of each motor so that we could start looking for appropriate drone motors. Our team also decided that we needed at least a 2-to-1 ratio between weight and thrust force. Given that the overall weight of the drone was about 2kg, divided by 4 it gives 500 grams per motor, and for a 2:1 ratio, we determined that we needed motors that could pull 1,000 grams or more.

Fernando was able to find 3 different motors and their recommended propeller diameters. Their specifications are the following:

The team also determined to use a 3-cell 5000mAh Li-Po battery since they were fit for our budget, had a good amperage, and did not exceed 400 grams in weight.

In order to decide which motor would be best for our project, Fernando ran multiple simulations on eCalc, which is the most reliable simulation tool for a quadcopter.
The main results are represented in graphical form in the images below:

The data collected from the simulations has led our team to select the iFlight XING X2806.5 motor.

It has the longest mixed flight time with or without payload that greatly exceeds the 4-minute mark set forth by the competition. Additionally, the motor has the highest efficiency in the no payload simulations and it has the 2nd best efficiency in the payload scenarios. 

With all UAV components finalized, I was able to design the UAV frame with some help from Fernando. Below is the final design, along with the UGV attached. 

Below are some videos of our final product, as well as some fly time.