Received October 19, 2017, accepted November 23, 2017, date of publication December 6, 2017, date of current version February 14, 2018. Digital Object Identifier 10.1109/ACCESS.2017.2780082 Human Expertise in Mobile Robot Navigation MOHAMMED FAISAL 1,2 , (Member, IEEE), MOHAMMED ALGABRI 2 , BENCHERIF MOHAMED ABDELKADER 2 , AND HABIB DHAHRI 1 , AND MOHAMAD MAHMOUD AL RAHHAL 1 , (Member, IEEE) 1 College of Computer and Information Sciences, King Saud University, Al Muzahimiyah 19676, Saudi Arabia 2 Center of Smart Robotics Research, College of Computer and Information Sciences, King Saud University, Riyadh 11543, Saudi Arabia Corresponding author: Mohammed Faisal (mfasial@ksu.edu.sa) This work was supported by the Deanship of Scientific Research at King Saud University through the Research Group under Grant RG-1438-071. ABSTRACT Numerous applications, such as material handling, manufacturing, security, and automated transportation systems, use mobile robots. Autonomous navigation remains one of the primary challenges of the mobile robot industry; many new control algorithms have been recently developed that aim to overcome this challenge. These algorithms are primarily related by their adoption of new strategies for avoiding obstacles and minimizing the travel time to a target along an optimal path. In this paper, we introduce four different navigation systems for an autonomous mobile robot (PowerBot) and compare them. The four systems are based on a fuzzy logic controller (FLC). The FLC of one system is tuned by an inexperienced human (naive), while the three other FLCs are optimized through a genetic algorithm (GA), particle swarm optimization (PSO), and a human expert. We hope the comparison answers the question of which is the best controller. In other words, ‘‘who can win?,’’ the naive, the GA, the PSO, or the expert, in fine tuning the membership functions of the navigation and obstacle avoidance behavior of the mobile robot? To answer this question, we used four different techniques for optimization (the naive FLC, GA, PSO, and FLC-expert) and used many criteria for comparison, whereas other research papers have dealt with two techniques at a time. INDEX TERMS Mobile robot, genetic algorithm, partial swarm optimization, fuzzy logic control, robot navigation, avoid obstacles. I. INTRODUCTION Researchers expect that mobile robots will be responsible for several tasks in human life. Examples include warehouse management, packet distribution and arrangement, product handling inside stockrooms, and in working in accessible but dangerous sites [1]–[3]. Navigation is definitively one of the strategic tasks for mobile robots. Many advanced approaches have been used for autonomous navigation; however, this subject has not been thoroughly elucidated to date [4]. Diverse formulations have been developed for the autonomous mobile robots navigation, during the last decades, moreover, these tremendous developments could not cope with the new robotic challenges that are becom- ing more challenging. These new types of situations are essentially owed to the dynamic and incomplete knowledge about the new complex and unknown environments. The diverse self-control techniques, such as fuzzy logic, neural networks, genetic algorithms, have widely been used to tackle this type of dynamic and compensate for some unknown knowledge [5]. Cao and al. used multiple types of sensors (sonars and cameras) and stored maps with a fuzzy logic navigation approach for the mobile robot [6]. Unfortunately, authors did not show all the simulation conditions and the experimental design in their paper, making it difficult to repeat their exper- iments, in a controlled environment. Another FLC approach for indoor navigation developed by [7], where authors used a fuzzy logic controller for tracking a target and controlling their Wheeled Mobile Robot (WMR). The authors concen- trated their focus on the robot navigation, without interest to avoid obstacles; they used the FLC uniquely to control the motion of the WMR. Faisal [8] have developed an online navigation system for their WMR (Scout II robot), including two FLC, within an unknown environment. A tracking Fuzzy Logic Controller (TFLC) is used for navigate control, and an Obstacle Avoiding Fuzzy Logic Controller (OAFLC) is used for obstacles avoiding. An indoor FLC system for an autonomous vehicle was presented in [9], where authors used a camera sensor and guided the robot to its goal through FLC. Nevertheless, the authors concentrated on navigation without 1694 2169-3536  2017 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. VOLUME 6, 2018