Effects of making unknown information explicit in computer-supported collaborative information retrieval. An empiric study in an urban search and rescue setting.

Abstract

When actors have to perform tasks in a team, whether the actors are humans or robots, communication and cooperation is of utter importance. Adequate information sharing and handling by team members can increase the performance of the whole team. A contemporary field in which this is important is urban search and rescue (USAR). Currently, in USAR situations the role of tele-operated robots increases in importance. Robot operators are able to share information in a common operational picture in order to create situation awareness. In this common operational picture, every team member can contribute instead of only one. This thesis describes a between-subjects experiment to test a specific functionality, the Explicit Unknowns functionality, within such a common operational picture that enables the users to make relevant, unknown information explicit. In other words, users can indicate which information they need to know but cannot provide themselves. Team members with the right capabilities are made aware of this unknown information and could improve the shared situation awareness by answering these unknowns, improving team performance in turn. Expected was that the use of this functionality would improve the efficiency of rescuing victims, cooperation, situation awareness and subjective performance. Measures were taken by individual questionnaires and registrations in the common operational display. An experiment was conducted using a simulated environment in which participants were paired as a team. Each team member had different capabilities and teams were given the task to fully register as much found victims as possible in the common operational picture. Certain victims required specific information that could only be provided by one of the team members. These cases required cooperation between team members, which is why completeness was essential in the task. The explicit unknowns would enable team members to work more efficiently and hence perform better as a team. Results, however, did not show significant effects between conditions. Nevertheless, tendencies were noticeable and seem most promising for cooperation, subjective performance and search strategy. While not significant, these tendencies did indicate slight improvements when the Explicit Unknowns functionality was available. Subjective feedback from participants also indicated a positive tendency towards the used functionality. Therefore, the main conclusion in this research is that making unknown information explicit in collaboration tasks might yield effects for some topics, although it is essential to make changes in experiment setup and team formation.