In addition to our main mission of participating in various UAV competitions around the world, we also pursue some projects on the side. These projects are done to help solve challenges in future competitions, create break-through technology or simply just for fun

Frequently, we take on brand-new cutting-edge projects, concepts and ideas. These are kept secret, not released to the public and often announced only after they are completed! Remember to check back once in a while to see if anything new and exciting has been created by us. In the meantime, below is a selection of projects we have worked on so far.

Project status: complete

We have recently partnered with GRASP lab to create a NSF competition at Penn. In preparation for this competition, we have created a new quadcopter capable of lifting a large payload, has front and down facing cameras for navigation, and is powered by an Intel NUC and Pixhawk.

Project status: active

One of the most crucial components of any of our ventures in autonomous robotics is the autopilot system. Over the course of the upcoming semester, we will be programming our own Autopilot based on the Pixhawk and ArduPilot hardware. This autopilot will be custom designed to suit our needs, including extensive modularity and easy-to-add plugins. Once we have tested our base code, various spin-off projects will be launched, current plans include optical recognition and navigation, autonomous acrobatics and search-and-rescue.

Project status: CAD complete; cancelled during manufacturing due to newer design

In this project, students are designing a custom aircraft that will subsequently be produced and flown both manually and as an UAV using our autopilot. The project is currently in the computer design stage, where different body, wing and empennage styles are being modelled. Once this stage is complete, the aircraft characteristics will be simulated in both CAD software and a flight simulator. Then, specific components will be tested in a wind tunnel. Finally, the entire aircraft will be manufactured using laser-cutting, machining, 3D printing, injection and molding.

Some specifications of this aircraft include a 2.5 meter wingspan, glider-like performance, easy-to-fly characteristics, pusher-prop configuration and a large payload capacity. The aircraft will be used in future projects, such as FPV and long-range flight.

Project status: completed

The aim of this project is to designing our own quadcopter from scratch. This includes designing and manufacturing the frame, calculating specifications such as weight, lift, power, etc. and figuring out which motors, propellers and ESCs will be used. The idea behind this project is to create a custom quadcopter that can be modified in future projects for specific tasks and that can serve as a test-bench platform for the autopilot project.

Our first prototype had a successful test flight in April 2015. We are designing new frames this semester for much larger sizes, better durability and improved utility. In the long-term, we plan to add more enhancements planned for a future spinoff, including using variable pitch propellers that would give us much better responsiveness and maneuverability.

Project status: completed Dec. 2014

The aim of this project was to take a manual radio-controlled aircraft and convert it into an unmanned autonomous vehicle. To do this, we used an open-source autopilot called the ArduPilot. The aircraft was also fitted with a wireless video link to a ground station and a telemetery data link to a laptop computer. As a result, we were able to fully control the aircraft from our laptop (still keeping a radio as backup though!) and achieved fully autonomous flight. Thanks to team's precise tuning of PID values, the aircraft was able to fly waypoint to waypoint, fly in tight circles around the home waypoint and successfully fight strong, turbulent and gusty winds.

Thanks to the success of this project, we know have a test-bench platform on which we can test our own autopilot software and our custom aircraft designs. In the future, we can also imporve these results by setting cross-track values to force the aircraft to follow the pre-planned route rather than flying waypoint to waypoint and testing this platform over longer distances.