The SiFA-based model for design systems provides a very powerful tool to build and also study design systems.
The model allows a designer to build a system that will consider many different points of view while designing an artifact, but the designer does not have to predict all possible conflicts between these points of view and resolve them while building the design system. The designer needs to provide the design knowledge from every point of view, and the system itself will take care of the conflicts.
This way of building design systems removes the burden of checking rule bases for consistency and also makes maintenance much easier, since the knowledge in different agents does not have to be conflict free. The separation of conflict knowledge from the design knowledge in the agents also has very important consequences in terms of the maintainability and the understandability of a system.
The main disadvantage of the SiFA model from a practical perspective is that the fine grained agent size is less efficient than using larger agents, but the understanding gained by SiFA research will show what functionality should be grouped together to produce more efficient systems.
Apart from the practical issues of having the SiFA model for building design systems, the framework proposed in the thesis also provides a new perspective to study design systems, multi-agent systems, conflicts and negotiations.
SiFAs have proved to be very helpful in gaining insight into many design related research issues and they have revealed new research paths that should be explored further.