The objective of this research is to develop a wearable prototype system to assist people doing outdoor running workout safely and effectively. One essential research issue is to correctly recognize a runner's state during a running workout process by analyzing contextual data obtained from sensors and GPS positioning device carried on by the runner. The running workout process is represented as a state transition diagram using the Space-Oriented Model. The state recognition is based on the state correlations with the runner's heartbeat rate and running speed. Our test results show that by utilizing the runner's state correlation, is more precise to recognize a runner's state as compared to the state judgment which is only based on detecting whether a sensed value exceeds some medical threshold value. Another function offered by the system is to provide the user an outdoor running course and may adjust the course according to the runner' speed, temperature, and course distance so as to advise the runner how to effectively burn the target amount of calories and safely achieve the exercise goal.