Abstract
In the modern world, adaptability is a key performance index for a group to be competitive. The use of well-designed computer tools by its members can help increase its adaptability. The purpose of Adaptive Collaboration (AC) is to obtain the optimal group performance, e. g., the maximum sum of the individual performances of group members in their positions for a concerned period of time. AC is a natural extension of Role-Based Collaboration (RBC). Our early model E-CARGO (Environments - Classes, Agents, Roles, Groups, and Objects) of RBC can become a fundamental model for AC. This paper points out the key issues facing AC researchers, analyzes the key factors related with adaptability, clarifies the AC process, and outlines some preliminary yet promising solutions to some issues.