The Intelligent Ground Vehicle Competition is an autonomous challenge that we have participated in since 2006.
Competition
The goal of this competition is to produce a robot that can navigate fully autonomously in specific environments while avoiding objects and executing path planning.
See the IGVC-2008 page for a discussion on our next robot entry and planning relevant to this year's IGVC. Also look at the upcoming IGVC Research Page for research done on past competitions years.
Vehicle Constraints / Qualification Rules
In order to enter the contest, each team must meet certain qualification criteria. Below is a summary of the constraints. Full details can be found at the official IGVC Rules page.
Dimensional Requirements
Length: Minimum 3 feet (0.914 m), Maximum 7 feet (2.134 m)
Width: Minimum 2 feet (0.610 m), Maximum 5 feet (1.524 m)
Height: No minimum, Maximum 6 feet (1.829 m) excluding antenna
Velocity
Vehicle speed must be hardware-governed to not exceed 5mph. This is determined with a time trial during qualifications. If the average speed exceeds 5mph the vehicle will not qualify. In addition, if the vehicle ever goes higher than the time trial speed, that run will be discounted.
Autonomous Functionality
The robot must be able to
- Follow lanes
- Avoid obstacles
- Navigate to GPS waypoints (within 2 meters)
Payload
Must be able to carry a 20 pound load with dimensions of 18" x 8" x 8".
Safety
Both mechanical and electronic emergency stops must be implemented.
- Mechanical e-stop button must be placed in the rear center of the robot at least 2 feet and a maximum of 4 feet off the ground level.
- Electronic e-stop (a radio device of some sort) must be able to stop the robot's motion for at least 50 feet distance.
Autonomous Challenge Competition
Design Competition
Navigation Challenge
The Navigation Challenge tests an entries ability to navigate to several GPS waypoints. Points are awarded for the number of waypoints traveled to in the shortest period of time. To get the highest possible raw score an entry must visit all waypoints and return to the start location without leaving the course or hitting any obstacles. Additional points are award as a tie breaker to the fastest time.
In 2007 the Navigation Challenge took place in a grassy field that was approximately 40m by 40m square(Need to double check distance). The GPS waypoints were split up into two groups seperated by a fence made of orange construction tape. Some oddities of the selected area was that one set of waypoints were on the side of a rather steep hill, it appeared more than a 7% grade IMO, that could appear as a large impassable region given some obstacle avoidance algorithms and sensor layouts.
JAUS Challenge
The JAUS challenges is primarily designed to generate interest around JAUS. There are two levels to the challenge which test the extent to which participants have implemented the JAUS protocol. The award for passing the JAUS challenge is $500 which will only awarded once per level. If you participate in both levels in a year you still will only be award $500. You may not compete again for the same level in following years once you have been award the prize money for that level. We have been award prize money for level 1 so in 2008 we plan to compete in level 2.
Level 1
Tests the warning system and if your robot can be signaled to start and stop some autonomous action using JAUS.
Level 2
Tests the ability to command your robot to move to a GPS waypoint using the JAUS protocol. Additionally tests the ability to query your robot's current GPS location.
Software
To find the software that has been used at IGVC take a look at the SVN repository in the directory /projects/IGVC/ where code is then separated out by year.
Sponsors/Grants
- Northrop Grumman - pending
- International Business Machine - pending
- General Motors - pending
We need $5,000 to cover the expenses for parts for one unmanned ground vehicles.
We want $25,000 to fund future programs and fully supply multiple robots for future competitions.
In the meantime we are setting up a procedure to generate the needed funds.

