The goal of an Intelligent Tutoring System (ITS) is to improve training efficiency by monitoring student performance and providing automated tutoring advice with the goal of increasing student learning and throughput. Traditional ITS development has focused on static problems, such as math and physics (Koedinger, Anderson, Hadley, & Mark, 1997; VanLehn, et al., 2005). Recent systems have targeted dynamic environments, such as Navy shipboard damage control and basic shiphandling maneuvers (Iseli, Koenig, Lee & Wainess, 2010; Peters, Bratt & Kirschenbaum, 2011). The research described here examines the advanced shiphandling task of mooring to a pier, which is the graduation exercise at the Surface Warfare Officers School (SWOS). To develop an ITS for mooring, many variables were considered, including ownship parameters (e.g., engine and rudder status), predicted future paths, and student behavior (e.g., number of orders, gaze direction). This development process involved creating and vetting a task analysis with SWOS subject matter experts (SMEs) and several iterations of system prototype testing. An effectiveness evaluation of the prototype was conducted with twenty novice shiphandling students at SWOS, split into groups that received either human or ITS tutoring only for a mooring to a pier scenario. Afterward, all students completed another mooring scenario without any tutoring. Across both runs, performance was evaluated using ship parameters, student behavior, and instructor scoring metrics. Analyzing a wide variety of performance measures showed no differences between the two groups, suggesting that the ITS was able to tutor as effectively as human instructors. Future work will involve developing additional advanced shiphandling scenarios and examining how the student-to-teacher ratio can be increased using a combination of ITS tutoring and instructor supervision.
Developing and Evaluating an Intelligent Tutoring System for Advanced Shiphandling
10 Views