Swarms Robots Navigation with Target Position Aulia Rahman Thoharsin Faculty of Computer Science, Sriwijaya University Robotic and Control Research Lab Inderalaya, Indonesia auliarahman.thoharsin@gmail.com Siti Nurmaini Faculty of Computer Science, Sriwijaya University Robotic and Control Research Lab Inderalaya, Indonesia siti_nurmaini@unsri.ac.id Abstract —This paper describes how swarm robots communicate with each other for achieving a determined target based on combination Fuzzy and Particle Swarm Optimization (PSO) technique. In the implementation fuzzy logic for robot navigation system and PSO as a search system for finding the best position. In this experiment three identical robots with some sensors are used, such as three infrared sensors, one compass sensor and one XBee. For determining the position of each robot, a camera with color detection method is utilized. The robot and the camera sensors are connected to a computer as an information center. The results shows that, swarm robots is able to determine who is the leader and has the ability to search the best pathto reach the target. Keywords: Navigation, swarm robots, Fuzzy-PSO I. INTRODUCTION Robotics system is relatively new and interesing field that has a potential to affect significantly the properties of engineering and science education at all levels. A recent development in robotics system is swarm robots, with the use of large and simple group will make a lot of profit at a speed time of target discovering bothof in a dynamic or real environment. The advantages of using swarm robots involve an improvement in dynamic range and more fault tolerance including an autonomous searching, rescue operations, autonomous decentralized system for protection and damage control. Currently there are many researches performed about the swarm robots especially on how to achieve the target. The target which is achieved by the swarm robot in case of is dangerous for human. The previous studies show that, there are at least three stages to find the target, such as searching process, tracking toward the target, and declaring a discovery targets [1,2]. For finding the target, swarm robots must be equipped some sensors to detect a particular target [1]. If one robot reach the target, it must declare the position by providing information such as the estimation of target position to other robots. It can save the time of othe robot movement in the searching process. Furthermore, the movement of other robots must be able to navigate in several of unknown environments [2]. In this experiment for solving such problem in swarm robots especially target seeking behavior, Fuzzy-PSO is utilized. II. RELATED WORK A system design is made to guide the movement of the swarm in order to not to interfere or collide with another object. We need a system that is commonly known as a navigation system. One of navigation system for the robot in order to navigate well is fuzzy logic controller method. The use of fuzzy logic is based on the ability to solve the uncertainty problems. The uncertainty problems can be found such as in actuators, and the environment. [4] In order to achive a detemined target, swarm robots need a control for guiding the robot towards the target. One of the methods that can be applied for robots control to search the target is particle swarm optimization (PSO) algorithm. This algorithm is based on the mathematical operations with primitive computational and not complex in terms of speed processing and memory utilization. The algorithm is very simple, and paradigms can be implemented in just a few lines of code. [5] This method is imitating the ability of animals looking for food sources. Each individual in PSO will be considered as a particle, [3] in case of swarm robots, robot represents the particles and the target position represents the available of food sources. III. EXPERIMENTAL SETUP A. Fuzzy-PSO Algorithm Swarm robots can move individually or in group. Each robot will search for the best position that can be achieved and face different circumstances. Thus in this condition, each robot must be able to move individually. But when one of the robots reaches its best position, the robots will move in groups. When the robot moves individually, we use fuzzy logic controller algorithm and when the robot moving in groups, the robot will move by use fuzzy-PSO algorithm. Therefore, we need to modify these two methods before we apply to the robot. This modification is done by giving a sign that the environment around the robot has an obstacle or not. To learn about the environment around the robot has any obstacle or not, then the fuzzy logic algorithm will be applied first. And the modification situate in the rule base of fuzzy, which at this stage of the algorithm determines how the state of the obstacles [6,7]. Figure 1 shows the flowchart of the fuzzy-PSO.