Automatic Task Generation and View Control for 3D Graphical Manual

3D graphics and animation have been used increasingly to provide on-line help and training of complex electro-mechanical devices. A good training system must be able to demonstrate clearly the correct dynamic sequences of using the device based on users’ current state and requested action. The system must also be able to provide optimal viewing angles and lightings for users to view the dynamic sequences. This paper describes the design of a 3D graphical manual system. The system uses a framebased knowledge structure to model the relationships and dependencies between the components of the device. The knowledge structure is sued to determine the sequences of animation to be performed in order to demonstrate a requested action. The system employs a viewing sphere technique to determine the best (dynamic) camera angles to view the progress of animation. It also adjusts the lighting condition dynamically to ensure that the main subject is adequately illuminated. A laser printer is modeled to demonstrate the feasibility of the system.