Algoritma Pendeteksi an Obstacle dan Furniture Menggunakan Metod E Triangulasi dan Scanning pada Robot Berkaki Enam yang Diterapkan pada Kontes Robot Pemadam API Indonesia

Abstract

Based on Kontes Robot Pemadam Api Indonesia (KRPAI) 2015 yesterday, the UKSW legged-division robot, R2C – Acheron failed to proceed to the elimination round in National tier because the robot crashed into the doll and pass through it, which mean disqualification. As the rule stated that robot which pass through the doll will disqualification. The seriously problem that make that happen is about the method to detect the doll. To detect the doll, the robot used sensor called GP2D12 that is an infrared sensor. The value of the sensor which is analog data but is only read 0 and 1 likely digital data. So the sensor is changed by infrared proximity sensor which is truly return digital data. Therefore the reading process from the sensor can be done quicker. The robot consists of two microcontrollers, sensors especially ultrasonic sensor and infrared proximity sensor, fire extinguisher hardware, and robot’s algorithm. Main microcontroller is functioned as main controller of system that’s used to give command to servo controller and fire extinguisher. The second one is functioned as side controller of system that’s used to read sensor’s data, process data, and then give the result to the main microcontroller. Sensors are used to help robot in navigation, movement, detect while line in the floor, detects fire spot, and detects the doll also detect the wall that can be used to wall following. Testing has been done in two sessions, which was to test robot’s detection and avoidance to the doll obstacle and furniture. In session one for detection and avoidance, the robot’s algorithm was tested 10 times. The doll had 3 candidate location which was tested for every different location and for 2 different doll’s material. So the total tested that must be done in session one was about 60 times. For detection the doll, the percentage success about 98.3% and the percentage of success for the robot to avoid doll was about 93.3% . The overall percentage robot to detection and avoiding the doll for about 91.67% . In the second session, the number of testing that must be done was about 120 times, which was tested in 6 different room’s configuration and 2 different location for the furniture’s configuration in the room. The tested must be done 10 times for every configuration. The percentage of success was 83.75% for robot to detect the furniture.