Prev-Talk | Next-Talk | All-Talks | Talks-Sorted
Talk-03: High Data Rate Vehicle Dynamic Testing with MOOS
Jordan Britt, Jameson Colbert, Auburn University
The ability of MOOS to log large quantities of high frequency data and any errors that can arise from this, is often an inquiry made by fellow vehicle dynamic researchers investigating MOOS. While there are tests demonstrating MOOS ability to handle large quantities of data, these are more often than not performed by sensors operating at lower frequencies, such as stereo vision. The proposed presentation would seek to present our work to help to better answer this question.
The specific tests performed were on a triple-trailer-semi-truck, where a host of sensors where distributed throughout the truck and three trailers, including GPS, IMU's, and CAN which accounted for a total of 12 MOOS processes running, 6 of which were operating at 100Hz or greater. In total over 200 different measurements were logged, using an Advantech PC operating on a 1GHz Celeron processor. This was made possible by making a minor alteration to the functionality of pLogger. Additionally, we discovered that using wild-card logging could tip the scales on our needed processing power, from which we discovered a way to potentially automatically generate missions files based on m_Comms.ServerRequest.
Hence, it is our desire to present working with MOOS for high data rates in a practical light, in which we demonstrate MOOS's ability to log high frequency data, potentially necessary changes to pLogger, and desired functionality changes or alternatives to wild-card logging.
DOWNLOAD the brief given at MOOS-DAWG'11.
Categories:
- MOOS Core
- High Volume Data Logging