In the field of design the pen-based system is a newly developed computer interface that provides the designer with the convenience of a pen in freehand sketches. But these pen-based systems only focus on an interface familiar to the designers and the application of the hardware and software that go with it, treating the pen only as a mouse-like input device. As pen and pad are devices for the pen-based system, the hope is that they can be endowed with more intelligent characteristics to let them interact with designeris gestures and become a creative source for the designers, while simultaneously preventing the design fixation encountered by designers during design process. This research utilizes the unintentional hand gestures made by designers, such as the designeris grip of the pen or movement involved in playing with the pen, putting it down, knocking it, twisting it or shaking it, during the thinking process or when running into a design fixation. From the interaction between the pen and the pad, certain actions may be generated to stimulate the designeris thinking process. This research uses a neural network as the main learning mechanism for the eventual development of a prototype of a pen-based drawing system that provides timely visual stimulation: a G Pen system.