Robot Intelligence Technology and Applications 4

In this paper we outline the approach of solving special type of navigation tasks for robotic systems, when a coalition of robots (agents) acts in the 2D environment, which can be modified by the actions, and share the same goal location. The latter is originally unreachable for some members of the coalition, but the common task still can be accomplished as the agents can assist each other (e.g., by modifying the environment). We call such tasks smart relocation tasks (as they cannot be solved by pure path planning methods) and study spatial and behavior interaction of robots while solving them. We use cognitive approach and introduce semiotic knowledge representation—sign world model which underlines behavioral planning methodology. Planning is viewed as a recursive search process in the hierarchical state-space induced by sings with path planning signs residing on the lowest level. Reaching this level triggers path planning which is accomplished by state-of-the-art grid-based planners focused on producing smooth paths (e.g., LIAN) and thus indirectly guarantying feasibility of that paths against agent’s dynamic constraints.

Intelligent behaviors generally mean actions showing their objectives and proper sequences. For robot, to complete a given task properly, an intelligent computational model is necessary. Recurrent Neural Network (RNN) is one of the plausible computational models because the RNN can learn from previous experiences and memorize those experiences represented by inner state within the RNN. There are other computational models like hidden Markov model (HMM) and Support Vector Machine, but they are absent of continuity and inner state. In this paper, we tested several intelligent capabilities of the RNN, especially for memorization and generalization even under kidnapped situations, by simulating mobile robot in the experiments.

This paper presents a hybrid Fuzzy-D*lite algorithm for smoothly navigating robots in an unknown terrain, in real-time. D*lite is a clever optimal, incremental and heuristic search algorithm that is known to be capable of achieving a speed up of one to two orders of magnitude over repeated A* searches. Given a target destination and an incomplete map, it is able to generate a sequence of waypoints for a robot, on the fly, performing course corrections whenever necessary, at a reduced computational time and memory footprint due to its incremental search capability. On the other hand, a cascade of fuzzy systems designed to take advantage of symmetry in the problem domain implements target pursuit and stationary spinning behaviours, for a two-wheeled robot. These reactionary systems calculate the exact steering angle and speed adjustments, enabling the robot to navigate smoothly and fast. We demonstrate how these complementary algorithms can be fused together to achieve smooth and fast continuous re-planning actions in a partially unknown terrain.

In order to make robots perform tasks autonomously, it is necessary for robots to know the surrounding environments. Therefore, a world modeling should be made in advance or concurrently. It is important to know an accurate position for the accurate world modeling. The aim of this paper is an accurate localization method for the world modeling under the situation where the portion of signals from global positioning system (GPS) satellites is blocked in urban environments. In this paper, we propose a detection method for non-line-of-sight satellites and a localization method using the GPS, the inertial measurement unit (IMU), the wheel encoder, and the laser range finder (LRF). To decide whether the signal from the satellite is blocked by the building, the local map that is made from the local sensors and an LRF is exploited. Then the GPS reliability is established adaptively in a non-line-of-sight situation. Through an extended Kalman filter (EKF) with the GPS reliability the final robot pose is estimated. To evaluate the performance of the proposed methods, the accuracy of the proposed method is analyzed using ground truth from Google maps. Experimental results demonstrate that the proposed method is suitable for the urban environments.

In this paper, we propose a simultaneous localization and mapping (SLAM) algorithm incorporating a dynamic switching mechanism to switch between FastSLAM 1.0 and 2.0, based on a threshold of effective sample size (ESS). By taking advantages of FastSLAM 1.0 and 2.0 through the proposed dynamic switching mechanism, execution efficiency is significantly improved while maintaining an acceptable accuracy of estimations. To show the effectiveness of our proposed approach in comparison to FastSLAM 1.0 and 2.0, several simulations are demonstrated in this paper.

Computationally efficient SLAM (CESLAM) has been proposed to solve simultaneous localization and mapping problem in real-time design. CESLAM first uses the landmark measurement with the maximum likelihood to update the particle states and then update their associated landmarks later. This improves the accuracy of localization and mapping by avoiding unnecessary comparisons. This paper describes a modified version of CESLAM called rapidly operations SLAM (ROSLAM) which improves the runtime even further. We present an empirical evaluation of ROSLAM in a simulated environment which shows that it speeds up previous well known algorithms by 100 %.

Computationally Efficient SLAM (CESLAM) was proposed to improve the accuracy and runtime efficiency of FastSLAM 1.0 and FastSLAM 2.0. This method adopts the landmark measurement with the maximum likelihood, where the particle state is updated before updating the landmark estimate. Also, CESLAM solves the problem of real-time performance. In this paper, a modified version of CESLAM, called adaptive computation SLAM (ACSLAM), as an adaptive SLAM enhances the localization and mapping accuracy along with better runtime performance. In an empirical evaluation in a rich environment, we show that ACSLAM runs about twice as fast as FastSLAM 2.0 and increases the accuracy of the location estimate by a factor of two.

The problem of development a sustainable efficient and accurate augmented reality system possessing the interactivity property and basically implementing markerless approach for supplementing the real scene is addressed in this paper. Custom implementation is based on applying of feature-based SLAM algorithm, which is resistant to change light conditions, able to track both textured and non-textured objects in real-time mode achieved by introducing a once-performed offline system teaching step. The proposed solution is universal because of the system’s configurability in accordance with the desirable result.

In this paper, a control strategy of RoboCup throw-in technical challenge for small-sized humanoid robot is proposed. Throw-in is one of the technical challenges of RoboCup competition, which is a well-known robot competition held annually in different countries. To complete this challenge, the robot must possess object detection ability, stable walking ability, ball-holding ability, and ball-throwing ability. Hence, a small-sized humanoid robot, aiRobots-V, is presented to accomplish the throw-in challenge. Moreover, the control strategy and the abilities mentioned above are introduced. The experiment results demonstrate the performance of the proposed method. Furthermore, aiRobots-V won the second place in this technical challenge of RoboCup soccer game.

The paper is concerned with control over a lower-limb exoskeleton device when it performs a sit-to-stand motion from a chair. A mathematical model describing full dynamics of the device is derived, and the strategies to facilitate the desired motion are outlined. A control system based on a modified Jacobian transpose is proposed, and its performance is evaluated in a series of numerical experiments. The simulations proved that the proposed control method can be successfully used to provide stable sit-to-stand motion from a chair.

In this paper, an adaptive tracking control of a directional drilling system subject to state delays and parameter uncertainties is investigated. The Explicit Force, Finitely Sharp, Zero Mass (EFFSZM) system is used to represent a directional drilling system. Invariance and Immersion (I&I) adaptive control approach is used to ensure the EFFSZM system to follow a set of predefined path autonomously. The performance of the proposed approach is compared to the one $$\mathcal {L}_1$$ adaptive control under the identical condition. To observe and validate the performance of the proposed approach, several simulation scenarios are conducted. The results show that both the controllers are able to stabilize the system in the presence of parameter uncertainties while maintaining the trajectory tracking error minimum.

Fuzzy logic systems have been implemented successfully for the design of a wide variety of control systems. They provide a powerful way for designing nonlinear controllers using human expert knowledge. In this article, we present an approach to design and implement a fuzzy logic proportional integral controller (Fuzzy-PI) for omnidirectional robot navigation system, using a field-programmable gate array (FPGA). First, we define the kinematic model of the robot system and then we design, simulate, and optimize the controller navigation system using MATLAB and Robotino Sim platforms. The main goal of this work is the design of the Fuzzy-PI controller and the hardware implementation using FPGA resources. The controller can be implemented on an FPGA using software or hardware approach. For the latter approach, the Fuzzy-PI algorithm is implemented in VHDL language, synthesized, optimized, placed and routed, and downloaded on an FPGA board.

This paper discusses the suitability of usage of smartphones running on Android operating system in robot navigation tasks. These phones are equipped with many sensors useful in robotics. Robot navigation tasks are often recommending high-performance devices (computers) to work on. Nowadays smartphones are also equipped with very powerful processors. Therefore a performance test of different devices with Android operating system including their sensorial equipment has been made to determine the usability of these devices for real-time robot navigation.

In the past decade, there have been a lot of researches on various models and controllers for two-wheeled inverted pendulum-type mobile robots. Two-wheeled inverted pendulum-type mobile robots are suitable for the service robots working both in the indoor and outdoor environments. In this paper, a bistable two-wheeled mobile robot which can tilt up and down (B-TMR) is designed and implemented. The B-TMR remedies the weak points of two-wheeled inverted pendulum-type mobile robots. The robot body is designed to maintain balancing state and tilted state. Cost efficiency is improved by using a single accelerometer with a Butterworth filter in order to measure inclined angle. To control five motion modes (Mode 1—Balancing; Mode 2—Tilt Up; Mode 3—Tilt Down; Mode 4—Tilted Move; Mode 5 - Balancing Move), a mode-selective controller is proposed. Through various experiments, the proposed hardware and controller are validated.

This article shows the advances on previous outcomes with a quadcopter in the context of Situated Robotics. Thus, external vision is used to capture the performance environment made up by four checkpoints over which the drone must autonomously fly. The implementation is discussed and the outcomes of the trials performed with the drone guided through two testing models—one external vision-centered and the other one hybrid—are shown.

We describe how control/status OO-messages on shared-memory middleware can provide better performing control of a micro-air vehicle (MAV). To illustrate this, we provide a new hardware abstraction for a controller application that is completely analogous to the popular ardrone_autonomy (AA) package that enables the Parrot AR Drone 2.0 quadcopter to be flown using commands over Wi-Fi. For fairness of comparison, we use the OO-messages on shared-memory middleware implementation gusimplewhiteboard in parallel with the ROSAA in the same code-base. We demonstrate the performance improvements associated with using gusimplewhiteboard messaging in place of ROSmessages and services. We explain how further performance improvements can be achieved by fully implementing the Time Triggered Architecture (TTA) of the gusimplewhiteboard and its associated tools (clfsm & LLFSMs).

An autonomous mobile robot has opened a new era of hospital logistics. Mobile robots will substitute human courier by transporting goods within the hospital day and night. Multiple robots will be deployed at a hospital and coordinating these robots will increase the efficiency of delivery. This paper introduces the framework of the simulator which enables testing coordination algorithms of multiple robots. The simulator models hospital environment with the hospital-specific delivery demand as well as the specification of robots used for the delivery. The outcome of the simulator is the performance of applied task allocation algorithms, which includes overall delivery time and waiting time of delivery tasks.

Social robots are expected to behave in a socially acceptable manner. They have to accommodate emotions in their decision-makings when dealing with people in social environments. In this paper, we present a novel emotion mechanism that influences decision making and behaviors through attention. We describe its implementation in a cognitive architecture and demonstrate its capability in a robot companion experiment. Results show that the robot can successfully bias its behaviors in order to make users happy. Our proposed emotion mechanism can be used in social robots to predict emotions and bias behaviors in order to improve their performances.

This paper introduces an approach to solve the task assignment problem for a large number of tasks and robots in an efficient time. This method reduces the size of the state space explored by partitioning the tasks to the number of robotic agents. The proposed method is divided into three stages: first the tasks are partitioned to the number of robots, then robots are being assigned to the clusters optimally, and finally a task assignment algorithm is executed individually at each cluster. Two methods are adopted to solve the task assignment at each cluster, a genetic algorithm and an imitation learning algorithm. To verify the performance of the proposed approach, several numerical simulations are performed. Our empirical evaluation shows that clustering leads to great savings in runtime (up to a factor of 50), while maintaining the quality of the solution.

In this research, we consider modeling of violin playing robot arm. MATLAB/SIMULINK are used for modeling of robot arm with seven degrees of freedom is considered, which is flexible than previously used robot arms with 6 joint for violin playing. Also, previous robot system which used Mitsubishi industrial robot arm (RV-2SD) is updated. In this model, torque, current consumption and voltage of each joint can be measured. Dynamixel-Pro from Robotis Co., Ltd. is used for the joints. This robot arm with seven joint has same range of movement (RoM) with human. This makes our designed robot good at violin playing. This paper presents basic violin playing technique, 3D modeling using Solidworks software, PID control system of servo motor using MATLAB/SimMechanics tool, and physical system of servo motor.

Robots can learn knowledge by observing demonstration of humans. As tutees, robots need to not only observe human behaviors, but also make proper feedbacks for human tutors because learning is an interactive process in which information is delivered in bidirectional ways between humans and robots. Gaze is an adequate method for robots to provide human tutors with feedbacks that robots are concentrating on current learning because gaze directly represents where they are paying attention to. This paper proposes a gaze control algorithm with a state machine in learning from demonstration. A human tutor shows demonstration in front of a robot tutee, and the robot tutee observes the demonstration for learning. The robot tutee perceives external environment through its camera, recognizes a human and objects, and figures out a state at which the robot tutee is situated. Then, the robot tutee gazes at proper targets that are predefined by the state machine. The human tutor also adjusts the demonstration to make learning more effectively according to the robot tutee’s feedbacks. The effectiveness of the proposed method is demonstrated through the experiments with a robotic head with 17 degrees of freedom, developed in the RIT Lab., KAIST.

This paper shows the design methodology of a humanoid robotic arm with realistic mechanical structure and performance. Realistic and mechanically robust structure for a prosthetic arm was developed in Solid Works. The torque, power requirements and cost estimation were assessed systematically by interfacing the model with SimMechanics software. The robotic arm is equipped with several robot servo motors which perform as links between arms and perform arm movements by interfacing with a robot servo controller and the PIC16F886 microcontroller. The robot servo controller has the capability to drive the servo in controlled position, speed, and acceleration modes. Due to the complexity of the arm kinematics, machine-learning techniques, which rely less on precise mathematical analysis, are implemented. ANFIS is one such machine-learning technique which helps in decision-making and control of robotic arms. This paper implements a MATLAB-derived multilayered ANFIS controller using a PIC16F886 microcontroller as a supervisory control for a 6 DOF robotic arm. This type of robotic arm has many advantages such as simple structure, high flexibility, low energy consumption, quiet operation, and sensory feedback which make it a prosthetic arm with very high resemblance to a normal arm. A good tradeoff between cost and performance is achieved in order to meet the goal of less expensive and useful robotic arm for the disabled. The practically built arm is tested with predefined paths and random positional targets with in work space and results are shown to act satisfactorily.

The minimally invasive surgical operation of intramedullary nailing has prevailed as a standard technique for a definite stabilizing treatment in femoral shaft fractures, and reduction is a critical step before carrying out an invasive operation. However, there are some disadvantages about this technology, such as the frequent radiation exposure for the operator and patients. This paper describes a novel cuff-type robotic-assisted femoral shaft fracture repositioning system, and preliminary discusses the constitutive modeling of passive transverse mechanical properties of skeletal muscle. Skeletal traction and fragments alignment for femur fracture reduction can be performed separately, which are standard procedures for surgeons in clinical. Correspondingly, the robotic system consists of distraction device, cuff-type reduction unit and remote control system. Preliminary experiments based on artificial models have been done, and the results show the effectiveness of the proposed robotic-assisted system, and indicate the potential value for further study.

This paper observes interactions between a robot and pets, especially canines, and also investigates the feasibility of pet care robot (PCR) as an alternative for human in terms of pet caring based on animal–robot interaction (ARI) with experiments. The paper demonstrates the effectiveness of PCR for playing with pets, which is designed based on ARI, ethology, and user-centered approaches. The experiments were performed to observe the interactions in two conditions: when the dogs meet the robot at the first time and when the dogs already meet the robot for some times. The result summarizes that canines are able to show better interaction performances with robots through repetitive learning with their owners. Furthermore, this result also suggests the feasibility of PCR as an alternative for human pet sitter in terms of the playing function.

Deep neural networks have achieved state-of-the-art performance for a variety of pattern-recognition tasks. In particular, the deep convolutional neural network (CNN), which is composed of several convolutional layers with a nonlinear activation function, pooling layers, and fully connected layers or an optional global average pooling layer, has received significant attention and is widely used in computer vision. Some research is now replacing a top fully connected layer with global pooling to avoid overfitting in the fully connected layers and to achieve regularization. This replacement is very important because global pooling with additional convolutional layers can eliminate restrictions on the necessity for fixed-size or fixed-length input in the fully connected layers. In this paper, the top global pooling layer is focused on, which is used in place of the fully connected layer and creates a simple and effective pooling operation called random crop (RC) pooling. Additionally, how to attain regularization in the top RC pooling layer is discussed. RC pooling randomly crops the feature maps so that only the images with sufficiently scaled and centered objects can be well-trained. This approach achieves comparable accuracy on the CIFAR-10/100 and MNIST.

Robotics projects of novice engineering students commonly focus on modeling predetermined reactive behaviors. This paper proposes an alternative approach, namely engaging students in creation of and experimentation with learning robots. To verify the feasibility of such approach, we conducted a case study, in which two novice engineering students constructed a humanoid robot and implemented a robot learning experiment. The project assignment was to build a humanoid robot capable to learn to adapt its posture while lifting various weights. The students successfully performed the project. Their robot learned from successes and failures of its trials while referring to analytical analysis made by a remote computer. Our case study showed that practice in teaching a robot to learn had significant advantages: It introduced the students to advanced concepts of robotics and AI, taught to perform engineering experiments by combining empirical and analytical methods, and inspired thinking about learning and meaning-making.

Loop closure detection plays an important role in vSLAM for building and updating maps of the surrounding environment. An efficient vSLAM system needs an informative descriptor for landmark description and stable model for making decisions. Most of the solutions dependent on using a single descriptor for landmark description, whereas other solutions proposed to use a combination of descriptors. However, these solutions still have the limitation in correctly detecting a previously visited landmark. In this paper, an ensemble of loop closure detection is proposed using Bayesian filter models for making decisions. In this approach, a set of different keypoint descriptors is used as input to bag-of-word descriptors. After that, these descriptors, i.e., SIFT, SURF, and ORB, are used to construct Bayesian filter models and ensemble learning algorithm for loop closure detection. The proposed approach is validated on a public dataset, namely City-Center dataset (CiC). The results shown that the proposed ensemble algorithm outperforms single model and existing loop closure detection system approaches. It gives 87.96 % for ensemble learning and 86.36 % for the best single model and 37, 80, 81 % for FAB-MAP, PIRF-Nav2.0, and RTAB-MAP, respectively.

In this paper, we advance the thesis that educational robots can make a much larger contribution in the classroom than has hitherto been the case. However, to realise this potential, it is necessary to supply teachers with detailed guides for classroom robotic student activities for achieving specific curricular learning objectives. As illustration, we describe in detail three activities that use the capabilities of the inexpensive smartphone-based mobile robot Zorro developed by us.

The goal of this paper was to accomplish a technical challenge of double passing soccer game for humanoid soccer robots in RoboCup competition. Using only a vision sensor, the control strategies for the technical challenges of humanoid league in RoboCup are designed and presented. The vision system includes the color space setting, the object recognition, a simplified mean shift algorithm, and the target position derivation. Vision system works on the tasks of object recognition, which includes the goal, landmark poles, and the interval of two black poles. The computational time is reduced greatly by the mean shift algorithm and that time can be utilized to do other control strategies. With the proposed control strategies, humanoid robots can successfully complete the RoboCup double passing task. The successful experiment results demonstrate the feasibility and effectiveness of the proposed foot–eye coordination control scheme.

We address the motion management system to apply it to the robot soccer game. The soccer system has a vision system, a host computer which manages the strategies and position control of the soccer robots, and a communication system between a host computer and three soccer robots. The player robots can take 32 motions according to the game situation. To avoid the obstacles such as the opponent player robots, we proposed the novel motion control method which consists of CM and PL-AMS. The CM decides the avoid direction considering the position and size of the obstacles. PL-AMS selects the appropriate avoidance motion by the fuzzy logic. In addition, we have the strategy for the game situation such as the offense and defense by using the seven subregion divisions. Moreover, to prevent the ping-pong effect, we provided the buffer region between subregions.

The development of mobile robots or in general autonomous mobile systems is a complex task. These systems are characterized by a set of sensors and actuators as well as software running on computing hardware. Depending on the foreseen tasks of designing a mobile system, the overall complexity of the design may range from a few integrated components to several tens of sensors and actuators including distributed information processing mechanism. Besides system modeling and simulations, the design process involves system testing by performing several experiments. A difficult part in this testing phase is to identify the relation between the observed system behavior and the underlying implemented hardware component or part of software. With our test environment and a specific validation method, an iterative design cycle can be enabled that helps to speed up the design process as well as achieve the system behavior that is defined by the specification.

Soft robotics is an expanding new research field. This article presents a state-of-the-art review on soft robotics including its research directions, key characteristics, materials, design and fabrication techniques. Although many biomimetic soft robots have been developed, a few of these have industrial applications. This article proposes a soft XY machine table for the purpose of object manipulation. The proposed table combines the concepts from three areas: soft robotics, object manipulation and industrial application. The surface of proposed table is entirely soft and embedded with inflatable air chambers. Surface deformation is generated by inflating these chambers. One object manipulation approach is to generate travelling waves on the deformable surface.

This paper presents a theoretical approach for a humanoid robot performing autonomously in a golf setting. The main concentration of this research is on using vision and mathematical calculations to perform and evaluate the success of each test. Most of the problems encountered in making a humanoid perform complex actions like playing golf arise from the constraints in their features and implementations. Parts of the solution comprise linear interpolation and angle calculation. Challenges faced and solved include some that will lead to a better understanding of human–robot interaction.

This paper describes the working of the system design for the Mars rover. The rover, developed to compete in the Mars Society’s University Rover Challenge 2015, was designed to perform various tasks such as site survey, sample return, equipment servicing, and astronaut assistance in a Mars-like landscape of dry, non-vegetated, rocky terrain. The complete design features a bioinspired eight-wheeled drive mechanism, an integrated robotic arm along with a stereo vision technique for advanced image processing. This paper focuses on the drive mechanism of the rover design. The 8-wheeled rover combines the rocker-bogie mechanism with four rocker wheels and four spider-leg wheels. The spider legs ensure that it can traverse over heights greater than the chassis height, which could be three times as much as the diameter of the wheels. NASA’s current rover can only traverse a height twice the diameter of the wheel. Additionally, the wheels are actuator-powered, and hence, the slope of the rover can be adjusted in such a way that it does not topple for a wide range of inclination allowing the rover to traverse over highly rugged terrain. The rover design can be modified for many applications notably the exploration of alien planets, deep sea trench, and other environments where human exploration is almost impossible. This effort to make the rover mechanism more efficient may one day be instrumental in detecting life and many such possibilities, in Mars and other planets.

A system of components designed for the rapid prototyping of different robot body plans for legged robots is described in this chapter. The components are designed and fabricated using 3-D printing technology. As such, they constitute a low-cost way to experiment with legged robots for researchers, practitioners, and students. As legged robots become more common, there will be a need for new designs and new intelligent behaviors. A system of components that facilitates experimentation can play an important role in this development. And because these components are 3-D printed, they can easily be shared and fabricated by others.

This article describes a rigid and soft combined robot that is designed to be used in confined spaces. The robot is based on pneumatically actuated peristaltic motion. This motion is accomplished by the use of a pair of pneumatic annular grippers in concert with a novel pneumatic soft actuator. The actuator is capable of actuation and it is made of soft material. So it can passively adapt to the outside passage, with this characteristic the robot can navigate through some constrained environments that would otherwise be inaccessible with conventional rigid actuators. The annular grippers consist of six scissor structures wrapped around the cylinder which can dramatically increase the maximum diameter of the grippers. So the robot can adapt its shape to external constraints and obstacles. The robot also contains a 2-DOF pneumatic soft joint that promises the robot capable of changing its direction actively.

A pneumatic soft robot hand which imitates the shapes and gestures of human hand is introduced in this article. Soft robotics, 3D printing technology, and casting technology are used to fabricate the soft robot hand. The robot hand consists of palm and fingers. The thumb has two joints, and other four fingers have three joints. The fingers are driven by soft actuators which are made of highly extensible, elastomeric material (Ecoflex 00-30; Smooth-On, Inc.). Kinematics model is established based on the homogeneous coordinate transformation matrix. The soft robot hand has good flexibility and compliance; thus, it can perform complex gestures and grasp objects with different shapes. The robot hand is mainly made of silicon rubber which makes it light in weight and compatible with different objects; additionally, it is even capable of lifting heavy objects. The heaviest object it can grasp weighs 0.55 kg. The pneumatic soft humanoid robot hand can be applied in industrial manufacturing and medical field.

The paper describes a concept of software tools for data stream processing. The tools can be used to implement parallel processing systems. Description of the task is presented in the first part of paper. The system is based on pipeline parallelism and was distributed for using on a cluster computer. The paper describes a base scheme and a main work algorithm of the system. An actual application example is presented. The system has some weak sides which are described at the end of paper. Direction of future research is presented at the end of the article.

Sensors have been developed and applied in a wide range of fields such as robotics and autonomous vehicle navigation (AVN). Due to the inability of a single sensor to fully sense its surroundings, multiple sensors based on individual specialties are commonly used in order to complement the shortcomings and enrich perception. However, it is challenging to integrate the heterogeneous types of sensory information and produce useful results. This research aims to achieve a high degree of accuracy with a minimum false-positive and false-negative rate for the sake of reliability and safety. This paper introduces a likelihood-based data fusion model, which integrates information from various sensors, maps it into the integrated data space and generates the solution considering all the information from the sensors. Two distinct sensors: an optical camera and a LIght Detection And Range (Lidar) sensor were used for the experiment. The experimental results showed the usefulness of the proposed model in comparison with single sensor outcomes.

Precast concrete (PC) is used in many construction sites around the world, which is delivered after it is totally made on the plant. This is because it can reduce the rate of structural defection and building time. While in construction, many of PC materials are moved by a crane operator to specific positions just by looking. And the workers use hand gestures to fit a shear pin on the lower beam with an upper part. In this paper, a vision-guided method that is more efficient and faster than the conventional PC assembly methods is proposed. A vision-based 6-DOF displacement estimation method measures the relative displacement between the camera located in an upper slab and the marker located at the shear pin on the lower beam. Usually, PC slab has a quite long width, and it is not easy to estimate a precise 6-DOF information with just a pair of a camera and a marker. To mitigate this problem, multiple pairs of cameras and markers can be configured for large PC member and the beam. This paper deals with 6-DOF displacement estimation by using vision-based localization with a planar maker for PC construction sites. A camera detects the corner of the planar marker at first, and sub-pixel information is obtained for the corner and then the data are transferred to a main computer via Bluetooth communication. The main computer calculates 6-DOF displacement with corresponding points in the world coordinate frame. To show feasibility and robustness of the proposed method, some experiments are performed with varying distances.

According to a recent research, illumination invariant features could be available to guarantee robust and reliable performance of vision systems. The research incorporated the imaging sensor properties to the vision algorithm, which is a general case in robotics contexts. However, these features are only applicable for the outdoor scenes. This paper develops an algorithm that enhances the features reported in the previous research. The features work for the indoor scenes with the proposed algorithm. Experiments are conducted to verify the performance of the proposed algorithm. It is verified that the features enhanced by the proposed algorithm become robust to illumination fluctuations.

Circle detection is one of the most important problems in image processing since circular objects, such as balls, are observed frequently in many natural and artificial environments and problems such as light variations, occlusions, shadows, circle-shaped objects, and real-time processing have to be managed. The previous methods are either edge-based which generally suffer from processing burden or color-based and cannot deal with color variations suitably. This paper presents a real-time ball detection framework that uses both color information and shape information together to detect and track a ball robustly. The results demonstrate superiority of the proposed method over the previous. After classifying the color space, image segmentation, building regions, and extracting objects, to reduce the computation, it is focused on finding the green horizon to eliminate a part of the image, which is beyond the field border. Afterward some filters based on circle-fitting methods, image moments are applied to detect and track the ball. The results show that these filters are real-time, robust against occlusion and are able to track the ball even for distances more than six meters.

Autonomous fixed-wing UAV landing based on differential GPS is now a mainstream providing reliable and precise landing. But the task still remains challenging when GPS availability is limited like for military UAVs. We discuss a solution of this problem based on computer vision and dot markings along stationary or makeshift runway. We focus our attempts on using infrared beacons along with narrow-band filter as promising way to mark any makeshift runway and utilize particle filtering to fuse both IMU and visual data. We believe that unlike many other vision-based methods, this solution is capable of tracking UAV position up to engines stop. System overview, algorithm description, and its evaluation on synthesized sequence along real recorded trajectory are presented.

We address the vision-based local obstacle avoidance method. The proposed method is based on the obstacle-dependent Gaussian potential field (ODG-PF). Basically, the potential-field (PF) algorithm is a global path planning algorithm. Therefore, we modify the conventional PF to apply the local path planning condition which does not need any environment information. The proposed method makes the robot to avoid the obstacles according to the decision of the optimal path by itself. In particular, we can apply it to the robot soccer system to avoid the collision with the opposite player robots.

This article is devoted to development of methods of positioning of unmanned aerial vehicles (UAV) with use of video cameras and methods of computer vision. UAV flies over the district, having a relief. Main feature of proposed method is using of Gabor’s wavelets to search of so-called reference points with known 3D-coordinates. Reference points are used from etalon images of the district (for example, satellite images) and looked for on the frame from the UAV’s camera. Also, comparison of feature points from the frame and a satellite image detected by SURF algorithm is used. The system of modeling of flight of the virtual UAV which allows testing the developed algorithms is realized.

Up to now, industrial robots with an open architecture have been applied to several tasks requiring high skills such as machining, sanding, and polishing. As one of the notable results, the open architecture allows system engineers at user side to design a feedback control system to develop robotic applications. However, porting of an application software into different makers’ industrial robots is not easy even though having an open architecture, because the specifications of interface between PC-based controllers and robots are not standardized yet. In this paper, easy transplantation of a robotic application is tried by using ORiN (Open Robot/Resource interface for the Network) middleware interface. The application is the trajectory-following controller with vibrational motion used in a machining robot. First, the removability of undesirable cusp marks on a flat surface is examined through machining experiments. Then, portability of the controller into a small-sized educational robot VE026A with ORiN SDK (Software Development Kit) is evaluated on a simulation environment called WINCAPS III. Finally, a 3D printer-like interface is proposed for the machining robot to be controlled based on STL data. The design, implementation, and experimental results are presented.

Google Glass, as a representative product of wearable Android devices, provides a new way of Human Machine Interaction. In this paper, we propose a method for the control of a Surveyor mobile robot via a Google Glass. In doing it, we establish Wi-Fi communication between the Google Glass and the mobile robot, wear the Google Glass to observe robot states and its surroundings, and navigate the mobile robot through head movements detected by rotation vector sensor mounted on the Google Glass. The method allows us to completely free hands in navigating the robot without need for a computer monitor. In order to demonstrate the flexibility of the proposed method, we control the robot to go through a maze in a simulated environment via the Google Glass.

With the growing need for indoor localisation solutions, this paper investigates the practical applications of wireless networking technologies based on the empirical study. By comparing between the two most widely used wireless technologies, aims to identify which technology, between Wi-fi and Bluetooth, is more capable in RSS-based localisation. Field experiments were conducted in order to collect the data to model the propagation of the two technologies. This study demonstrates that, through comparing the derived models to empirical data, Bluetooth has the potential to improve indoor localisation methods due to its more accurate model.

Despite an idea of robotic system teleoperation is a relatively old concept, here we present its enhancements heading to an interconnection of teleoperation and collecting relevant information from the environment where robots act. This environment should be an intelligent space featured with various devices and sensors, which allows to obtain, preprocess and stores data in the cloud. Those data should provide relevant information for teleoperator or directly for robots, which act autonomously. For this purpose, we developed cloud-based tools, named Telescope v2. It is a platform-independent system for remote monitoring and controlling various systems. In this paper, we introduce this system, its abilities, and compare it with its network-based ancestor, Telescope v1. We analyze measurements of latency and response time when our new system is used for teleoperation in different places equipped with various Internet bandwidths.