2002 Overview 2003 Overview 2004 Overview 2005-06 Overview 2007 Overview 2008 Overview

    2008 System Overview Under Construction



        System Components:
        Aircraft - Vision - Ground Station - PNI - 2005 Computer Vision System








Sig Rasscal 110

Named after a famous explorer of, and city from the developer's home state. JOLIET is being designed to add an extra step of autonomy into our system in the spirit of the goals set by the International Aerial Robotics Competition, by assisting in removing humans from much the planning components of a successful mission. The software is currently written to analyze geo-referenced aerial imagery and detect man-made structures. Upon successful analysis, the program provides a list of GPS positions that can be utilized in other software for autonomous path planning. Detection of roads and other features are planned for a later release when our software team gets some free time on their hands.

JUAV (Java Unmanned Aerial Vehicles)

(Screenshot Coming Soon)
JUAV is the software that ties our entire system together. Incredibly tedious description to follow soon.

Ground Station


The Piccolo Model Builder software was written to shorten the time to transition from design to simulation by our mechanical and aerospace engineers. Currently, in order to prepare a set of gains for use by our simulator - a lengthy process involving collecting parameters about the construction of the UAV being simulated is required. This is usually done by manually editing .txt files in a program like notepad, where small typos can go unnoticed, produce unanticipated results in the simulator, and cause a lot of stress for everyone. By making a graphical based editor, where parameters can be quickly found and entered. Building a simulation model from a design in Pro/E or Solidworks can be as easy as copy and paste. Another important feature is the tooltip system, where if an engineer is unclear about a specific parameter, they can mouse over the item and get help quoted from the official documentation without the hassle of looking it up. The Piccolo Model Builder was completed just in time for use with the newest aircraft to join our fleet, the Sig Rascal 110.

PNI (Piccolo Network Interface)



The simple display disguises how incredibly useful this piece of software is to our team.
The PNI is a piece of software written in C++ using the Piccolo Software Development kit that allows communication with the CloudCap Technologies Piccolo Autopilot via TCP/IP and our custom UNI Protocol written to control unmanned vehicles.

Vision System / Object Recognition (2005 System)



This system was written last year, but was not taken to competition due to difficulties facing our team at that time. It's project name was UAVision, which we all thought was clever at the time. The system was implemented using Intel's OpenCV library which provides for an extremely fast and well written Computer Vision API. The algorithms and approach were based off techniques developed by the University of Texas, Austin's IARC team, as described in their 2004 paper. This provided a very fast implementation that worked at near real-time speeds, which was a vast improvement over the system developed by previous members in the 2003 system - which on the same hardware could process 3 frames per second. Sadly, the 2006 system cannot be described here yet. We've got a few tricks up our sleeves for Georgia this year, and ideas that to the best of our knowledge haven't been seen yet in UAV systems both academic and commercial. Check back here for details to be leaked as we get closer to competition time!