Project Number:
07-09
Research Project:
Vision-based Autonomous Ground Vehicle Design Within a Mixed Environment
P.I. Name & Address:
Alice C. Parker
Department of Electrical Engineering
EEB 348, Mail Code 2562
University of Southern California
Los Angeles, CA 90089-2562
Phone: 213-740-4476
E-mail: parker@eve.usc.edu
Project Objective:
International trade is growing rapidly, and with this growth come challenges due to port congestion, along with concerns for human security and safety. Congestion due to the lack of inland ports and massive demands due to shipping volume have resulted in delays in processing time, inefficiencies in the use of port resources, and high cost. Air pollution due to idling vehicles also contributes to the challenges faced in the design of future port terminals.
Our objective is to provide a solution to the port challenges through increased use of technology. In particular, we believe a solution can be found in the Automated Terminal (AT) and the use of Autonomous Ground Vehicles (AGVs) in the Port of the Future. Autonomous (robotic) vehicles in particular are a focus of interest because appropriate use of such vehicles can alleviate congestion and air pollution in future port terminals while enhancing safety and security. AGVs can possess sophisticated software algorithms that deal with these issues.
In particular, we will focus on applying the knowledge we gained through participation in the DARPA Grand Challenge, along with vision research we have been performing, and research by others in the autonomous vehicle community, to the problem of autonomous ground vehicle vision in mixed (human and robot) environments. The environment that will be studied most closely is the port terminal environment.
Task Descriptions:
Survey and evaluate previous Grand Challenge vision systems
Survey and compare other vision systems for robotics
Survey and compare urban Grand Challenge vision systems
Collect video data for roads and terminal areas
Apply wavelet transforms to video images
Implement Kalman filtering for learning
Implement Genetic algorithm for learning
Implement neural network for learning
Compare learning algorithms
Produce final reports and website
Milestones, Dates:
September 1, 2006 – August 31, 2007
Total Budget:
$89,613
Student Involvement:
One student at 50% effort x 12 months
Relationship to Other Research Projects:
Related to 01-16, 00-16;part of goods movement and international trade focus area
Technology Transfer Activities:
Project report posted on the website
Potential Benefits of the Project:
More efficient goods movement; improved port operations, reduced emissions, reduced port delays, enhanced economic development of ports
TRB Keywords:
Automated terminal, autonomous ground vehicles
Primary Subject:
4b.5 R & D resource base
Goals:
4c.3 Economic growth and trade
Enabling Research:
4c.8 Computer, Information and Communication
Modal Orientation:
4c.14 Maritime
