Abstract :
The type of disease that contributes to the highest mortality rate in the NonCommunicable Diseases (NCD) group is cardiovascular disease, one of which is
stroke. Diseases such as stroke can be prevented before or after they occur again by
reducing risk factors, one of which is by practicing body movements. Therefore, a
static cycling virtual reality system was developed to improve body balance for
post-stroke sufferers. The development of static bicycle virtual reality games cannot
be separated from the need for Non Player Characters (NPC) that are present in the
form of autovehicles as trainers for players. The NPC as a trainer will guide the
player during the simulation process by adjusting the speed according to the player's
ability. Developing a complex AI algorithm, of course, takes a long time and
expensive resources. The problems in NPC speed modulation can be solved by
using machine learning, one of which is using fuzzy logic. The purpose of
implementing NPCs as Trainers in this study is to create a technology or machine
that can facilitate humans, this goal is obtained based on the conditions of this
modern era where technology has developed rapidly. Fuzzy logic in this study uses
two classes for input and one class for output whose training data are obtained from
players. Input distance and heartrate from the player can make the NPC adjust the
speed of the player. The test results show that the results of the NPC speed of Fuzzy
Logic are able to make players do exercises stably.