High risk of climbing coconut tree manually become the main reason to build coconut tree plucking robot, not only the abnormality of bone but also the risk of falling from the coconut tree. The coconut tree plucking robot is made with the hope for helping people to pluck coconuts at the coconut tree easily and safely. Coconut tree with its condition make the coconuts difficult to be detected using image processing. Previous methods which are only work in indoor, only detect a coconut and only work on nearly uniform background are not suitable and easy to be disturbed with the interferences from the real condition in a coconut tree. An image processing with particle swarm optimization (PSO) method is introduced in this paper. It will find the best position of the coconuts at the tree and pluck it by giving a command to the arm to move toward the coconuts and cut its base by turning the grinder on the top of arm. Experiment results show that successful rate of the method to detect coconuts at the tree with cluttered background is 80% and then pluck them using the robot arm.

In this study, the teleoperation robot control on a mobile robot with 2D SLAM and object localization using LPWAN is proposed. The mobile robot is a technology gaining popularity due to flexibility and robustness in a variety of terrains. In search and rescue activities, the mobile robots can be used to perform some missions, assist and preserve human life. However, teleoperation control becomes a challenging problem for this implementation. The robust wireless communication not only allows the operator to stay away from dangerous area, but also increases the mobility of the mobile robot itself. Most of teleoperation mobile robots use Wi-Fi having high-bandwidth, yet short communication range. LoRa as LPWAN, on the other hand, has much longer range but low-bandwidth communication speed. Therefore, the combination of them complements each other’s weaknesses. The use of a two-LoRa configuration also enhances the teleoperation capabilities. All information from the mobile robot can be sent to the PC controller in relatively fast enough for real-time SLAM implementation. Furthermore, the mobile robot is also capable of real-time object detection, localization, and transmitting images. Another problem of LoRa communication is a timeout. We apply timeout recovery algorithms to handle this issue, resulting in more stable data. All data have been confirmed by real-time trials and the proposed method can approach the Wi-Fi performance with a low waiting time or delay.

This paper presents a new 3D map building technique using a combination of 2D SLAM and 3D objects that can be implemented on relatively low-cost hardware in real-time. Recently, 3D visualization of the real world became increasingly important. In robotics, it is not only required for intelligent control, but also necessary for operators to provide intuitive visualization. SLAM is generally applied for this purpose, as it is considered a basic ability for truly autonomous robots. However, due to the increase in the amount of data, real-time processing is becoming a challenge. Therefore, in order to address this problem, we combine 2D data and 3D objects to create a new 3D map. The combination is simple yet robust based on rotation, translation, and clustering techniques. The proposed method was applied to a mobile robot system for indoor observation. The results show that real-time performance can be achieved by the system. Furthermore, we also combine high and low-bandwidth networks to deal with network problems that usually occur in wireless communication. Thus, robust wireless communication can be established, as it ensures that the missions can be continued even if the system loses the main network.

This paper describes several methods in the development of simultaneous localization and mapping (SLAM) for real-time implementation on low-power wide-area (LPWA) communication. In a case of harsh environments, a wireless communication network may be having damaged. Thus, long range (LoRa) is used as a sub-communication. We use an autonomous mobile robot and a PC server running with MATLAB software to visualize SLAM information through graphical user interface (GUI). The robot employs light detection and ranging (LiDAR), encoders, and an inertial measurement unit (IMU) sensor to observe the environmental conditions. Accuracy and computational cost are the two main factors in SLAM determining the success of the method. Thus, we also propose the optimization methods to achieve both accuracy and computational cost. The proposed methods have been implemented in real-time observation for indoor environment. The experimental results show that the proposed methods can be used to optimize SLAM in real-time implementation, especially for remote monitoring through wireless communication.

This paper presents an online topological mapping method on a quadcopter with fast-growing neural gas. Recently, perceiving the real world in 3D space has become increasingly important, and robotics is no exception. Quadcopters are the most common type of robot working in 3D space. The ability to perceive 3D space is even required in order to enable real-time autonomous control. Dense maps are simply unpractical, while sparse maps are not suitable due to a lack of appropriate information. Topological maps then offer a balance between computational cost and accuracy. One of the most well-known unsupervised learning methods for topological mapping is growing neural gas (GNG). Unfortunately, it is difficult to increase the learning speed due to the traditional iterative method. Consequently, we propose a new method for topological mapping, called simplified multi-scale batch-learning GNG, by applying a mini-batch strategy in the learning process. The proposed method has been implemented on a quadcopter for indoor mapping applications. In addition, the topological maps are also combined with the tracking data of the quadcopter to generate a new global map. The combination is simple yet robust, based on rotation and translation strategies. Thus, the quadcopter is able to run the algorithms in real-time and maintain its performance above 30 fps.

This paper presents a plane extraction method for embedded computers in mobile robots which typically have less power than general-desktop computers. The increasing demand on immersive technology, including VR, AR, MR, XR, has encouraged to perceive the real-world in 3D space, and so is robotics. However, a problem occurs when large amounts of data are generated, especially for the embedded computers that are usually installed in the mobile robots. Therefore, we propose the plane extraction method to compress the data into other forms suitable for the mobile robot applications. PSO-seeded region growing algorithm was applied to segment the data based on planar characteristics. The output is polygons of each detected plane that can be projected to represent the captured objects in 3D space. It is simple and easy to be applied to the embedded computers, hence less computational time. The algorithm was verified using a depth image dataset of our laboratory. Based on the experimental results, a maximum of 15.1 fps can be achieved.