e-learning market is in high development. As a result, the research is continually being for more effective education system.
Previous e-learning which consists of general theory lectures with videos mainly, has developed for experiential learning which provides skill training in virtual environments(e-training). It can elicit the interests from trainee, and also solve the safety problem about high risk training, cost, space and place, and equipment failure problems, so on.
e-training is required a fair quality of simulation realization for virtual environments construction. However, the model of simulator is complicate to be an object, or it is very difficult to realize the all functions about trainees’ unpredictable movements. Therefore, the range and size of simulator are reduced when it is developed, so that the reality and interest are decreased.
Previous research is only active for the 3D virtual reality from computer graphic technologies, augmented realities, and action recognitions other than approach to the mentioned issues.
As the fact that ‘open-world’ games like GTA or Mine-craft are recently has been popular, trainees are still prefer the one has more freedom than things which are not.
This paper proposes the simulations with inference model due to these freedoms. A simulator with an inference model is able to effectively react to trainees’ unpredictable orders.
In this paper, design the inference models for the car maintenance training, and apply them to develop the intelligent simulation agents. Construct knowledge bases using JESS, and design the inference models to use it for the simulator. Based on these designed inference models, develop the simulation agents to confirm if the simulation correctly operates.
As the final outcome, trying to show the applications of intelligent simulators with e-training, and the possibility it is able to be used in systems other than cars.