| Abstrakt: |
In this paper, we describe and evaluate an innovative visual workbench for efficient Root Cause Analysis (RCA) of complex problems by support teams. Problems in business-critical systems typically involve multiple modules, requiring various subject matter experts (SMEs) to work together. The SMEs are often geographically distributed, located in different time zones, and need to jointly resolve issues under time pressure. There are significant challenges in these scenarios, such as lack of shared understanding, fragmented collaboration and limited knowledge reuse. Text and voice-based communication as used today is not efficient for performing joint problem analysis by distributed teams. Interactive groups engaged in sharing of ideas suffer from cognitive interference as well. Overhearing and/or reading others' inputs in the group work interfere with the cognitive process of group members (Nijstad, Stroebe, and Lodewijkx, 2002). Our workbench is designed to alleviate these difficulties through a visual toolset for collaborative Root Cause Analysis that supports multiple SMEs working together in their individual areas of expertise. The workbench boosts shared understanding using a visual Cause Tree diagram that can be concurrently updated by multiple users. It also enables 'in-context' collaboration by allowing the annotation of individual causes with relevant information such notes, actions and decisions. The workbench also performs automatic case matching with prior solved problems and self-learns from user actions, generating intelligent recommendations to help solve problems faster. Use of cognitive maps in decision making was suggested by Eden (1988). Chen and Lee (2003) developed a prototype of cognitive decision support system for strategic decision making situations. We propose the use of visual cognitive aids in problem solving tasks. We propose that in business critical systems, the use of this visual collaborative RCA workbench will: • improve the shared understanding of SMEs engaged in solving problems. • reduce cognitive interference of SMEs engaged in sharing of ideas to solve problems. • decrease the mean time to resolve problems. • improve knowledge sharing effectiveness of the SMEs working collaboratively. We intend to conduct a "Posttest-Only control group design" laboratory experiment to test our propositions. The participants of the experiments will be graduate level university students who have undergone courses on information systems development and project management. Groups of five to seven members will be formed and each group will work on an experimental task which will relate to problem solving in business critical systems. Treatment groups will use a prototype of our RCA workbench while the members of the control groups will be engaged in text and voice based communications to work on the experimental task. We intend to use the validated questionnaire to measure shared understanding, cognitive interference, and knowledge sharing effectiveness while an objective measure will be used for MTTR. [ABSTRACT FROM AUTHOR] |
