| dc.description.abstract | Sensor is a device in a robot to learn things about its surroundings. One of the problems from using sensor is its accuracy when retrieving data. This study developed a fire fighting robot sensor system using ultrasonic range sensor, fire detector sensor, and line sensor. In the process of retrieving data, these sensors require some pre-processing steps in order to obtain data that is equal to the actual data, such as calibration and data smoothing. Data smoothing methods utilized in this study are moving average and simple exponential smoothing. These methods are used to smooth the received realtime data especially in the range sensor HC-SR04 data. The results show that moving average with 5 windows and exponential smoothing with smoothing factor 0.5 can be used to reduce the ripples due to the existence of irregular data. However, both methods may slow down the robot‟s decision making process. | en |
