The engineering design process evolves products by a collaborative synthesis of specifications, personnel and organizations. Unfortunately, collaborative effectiveness is thwarted by existing single-user computer-aided applications like computer-aided design, computer-aided analysis, and others. These applications and associated file management systems assign editing rights to one technical person, e.g., a designer, analyst, or a process planner. In the absence of collaborative computer-aided engineering applications, we conducted a survey to establish that product collaboration is limited to interactive, either formal or ad-hoc design sessions, social communication tools, serial model sharing, terminal/screen sharing, and to conference call interactions. Current computer-aided (CAx) tools do not permit simultaneous model changes by a collaborative team editing the same model. Although over a decade of prior research has demonstrated multi-user feasibility for computer-aided applications, the architectural breadth of this research has apparently not yet compelled developers and end-users to develop and adopt new multi-user computer-aided applications devoted to product development.

Why have collaborative engineering CAx tools not been commercialized for mainstream use? This paper uses several multi-user prototypes, including the first Computer-Aided Engineering multi-user prototype called CUBIT Connect, to expose additional architectural hurdles to implementing new multi-user collaborative paradigms. These challenges relate to variable algorithmic performance times, multi-threading and event driven client notification processes, distributed access level security, and model change management in design sessions.