Autonomous Unmanned Drone
(2018-2019)
Note that I currently still work with the team as an advisor, but have given up my title as Chief Engineer, as I have too many other things that demand my attention.
​
The scope of this page covers work from 2018-2019.
Background
I was the Chief Engineer for the University of Houston's Space City Drone Team during its first year.
​
We were designing for the AUVSI SUAS competition, an annual international competition for students to design and construct an unmanned autonomous drone.
​
The drone was tasked with flying a mission completely autonomously. This includes:
-
Taking off and landing
-
Flying waypoints while avoiding certain regions of space
-
Staying in strict flight boundaries
-
Recognizing and classifying ground targets via onboard camera
-
Airdropping an unmanned ground vehicle (UGV)
​
​
​
Team structure
Since we had a pretty complex task, it was critical to have good organization. The project manager for the team decided on the hierarchy seen to the right.
​
Each of the bottom 4 boxes is a subteam.
Each subteam:
-
Worked on a specific group of related problems
-
Had ~8 members (people like to join more than one)
-
Was led by a subteam leader, who assigned tasks
​
Our team held mandatory general meetings once a week. Subteams met to work on their tasks at their own discretion.
​
Our team had ~20 active members, but as a first-year team, we expected that number to grow in future years as we find our stride and are better able to utilize more people.
​
Example ground targets
Team structure
Subteam responsibilities
My role as Chief Engineer
As the Chief Engineer, I oversaw the mechanical aspects of the drone. This included the physical design of the aircraft as well as the UGV. I managed the subteams responsible for these aspects of the drone, which were: flight systems, payload, and some of avionics.
​
I did some project management in order to direct the mechanical focus of the project. The people in the mechanical subteams made my job as a leader easy - they were happy to work on the project, and although we had a few delays, we got everything done.
Although being a leader was a great experience, I loved working on the project myself. I was very interested in all facets of the engineering problem. I enjoy programming as much (or more) as I enjoy mechanical design, and I was happy to get more experience with electronics. Most of all, I enjoyed working with the talented members of the drone team and learning more about aerospace engineering.
​
OUR
PROGRESS
This focuses mostly on things I've directly contributed to.
​
Synthetic Image Generator
I wrote a script in MATLAB to generate training data for the vision subteam. They used a neural network to analyze pictures from the drone. My script can quickly generate thousands of images of targets, with varying shape and color.​ To the right are a few examples.
​
Test plane
Using some plans available online, the airframe subteam built a plane to test essential systems without risking the competition plane. Below is an image of the plane, and a video of its maiden flight in heavy wind.
​
​
​
_JPG.jpg)
RESULTS &
LESSONS LEARNED
How our team did
Unfortunately, due to a variety of problems, our team ended up withdrawing from the competition, so we never competed. This was due to a few things:​
-
Lack of funds​​
-
Undertested software
​
Due to some unforeseen circumstances and a few organizational missteps, we ended up with lower funds than we'd like at the end of the year. We could have afforded to get to competition, but we would have sent far less people than we wanted. Instead, we chose to put the remaining funds towards the following year's efforts.
​
We could have sent people, but we believed that our software stack was under-tested. Simply put, we didn't trust it to be performant, and definitely not competitive enough to earn us a top spot. We were also rushing to finish some components that wouldn't be done in time.
​
​
My takeaway
I think that the team was and is a great idea, and I've learned a lot about project planning and how to manage people. I've learned that engineering as a big group isn't very easy.
​
I've also learned a lot about planes and drones, and I've thoroughly enjoyed the software components I was able to work on.
​