GUI Techniques for Assessing Autonomous Vehicle Behavior

Period of Performance: 02/21/2001 - 02/21/2003

$750K

Phase 2 SBIR

Recipient Firm

Cybernet Systems Corp.
3741 Plaza Drive Array
Ann Arbor, MI 48108
Principal Investigator

Abstract

As autonomous vehicles journey down the path of increasing competence, it is necessary for humans to be able to "look over their shoulder" to help guide and correct them. However, the further down the path UAVs travel, the more complex they become. The more complex they become, the more difficult it is to determine why they do what they do. Furthermore, more complex UAVs will be capable of following more sophisticated missions. As these mission become more sophisticated, the effects of deviations from these missions will become more difficult to resolve. Human operators will need tools to help them quickly and robustly determine why a UAV has made the decisions it has made, as well as the effects of those decisions on the UAV's mission. By applying imagination and creativity tempered with the basic principles of human-computer interaction and scientific visualization, this proposal will show that innovative graphical user interface techniques can be developed to solve these problems. These techniques then can be used to extend an existing UAV GUI and help operators come to terms with the obstacles presented by increasingly sophisticated autonomous vehicles.