Collaborative Knowledge Based Engineering

More Info


Present Knowledge Based Engineering (KBE) applications do not facilitate collaborative processes due to lack of process formalization and orchestration capabilities, leading to ad-hoc solutions. This increases application development time, reduces re-usability, and increases the cognitive burden of users leading to decreased design efficiency. Generic solutions from Workflow Management (WfM) can capture process knowledge in models outside of the KBE application and enable better collaboration. However, a challenge is to maintain flexibility when using WfM, which is rigid compared to other groupware. The research work of this thesis contributes technologies for saving and accessing the information of a KBE application to enable a methodology that flexibly facilitates collaboration by taking advantage of runtime caching and dependency tracking to dynamically populate tasks based on runtime context. These contributions are highlighted through case studies that demonstrate how present limitations of KBE and WfM can be overcome to open a new frontier for the next-generation of engineering software. Based on a verified software prototype, the key research contributions of this thesis are four-fold: (a) development of an information modeling approach making use of GraphQL to flexibly query KBE models to support collaborative activities, (b) development of a generic persistence capability for ParaPy to support transactional usage and retain full design history, (c) implementation of the worklet concept in KBE to increase flexibility of workflows and provide new form of process automation, and (d) creation of correlated dependency technique which has potential for use on emergent workflows. These contributions are paramount to undertaking multiple opportunities for collaborative KBE applications. It is expected that these research contributions provide a theoretical basis for achieving improved collaborative usage of KBE applications.