Autonomous weeding robots offer significant potential for reducing labor and environmental costs in precision agriculture. However, most systems decouple chassis locomotion from manipulator control, limiting efficiency in unstructured environments. We present LSTM-PPO-Weeding, an end-to-end framework unifying mobility and manipulation via memory-augmented reinforcement learning (RL). A high-speed region of interest (ROI) filter abstracts camera input into three-dimensional weed/crop poses, enabling the system to focus solely on task-relevant features. To mitigate challenges arising from occlusion and partial observability, we integrate long short-term memory (LSTM) into the proximal policy optimization (PPO) algorithm, establishing the proposed LSTM-PPO framework. The agent learns coordinated actions through a custom reward function balancing speed and precision. Experiments in simulation and real-world greenhouse settings show that LSTM-PPO-Weeding improves throughput while maintaining high accuracy. Compared to baseline methods, our approach reduces average weeding time by up to 74% with minimal loss in success rate, demonstrating its robustness for agricultural mobile manipulation.

Robotic tool usage still falls short in fine-grained task execution, specifically, learning how to select tools and determine configurations, achieving task compatibility, and enhancing subsequent manipulation remains challenging. Thus, this paper proposes a Graph Attention Grasp (GATGrasp) concept, which is dedicated to establishing a task-aware affordance grasping framework. It integrates a Visual-Language Model (VLM) for tool localization and segmentation across diverse scenes, while employing the GraspNet component to generate 6D candidates. The approach is then devoted to leveraging multimodal semantic and geometric features, learning the spatial relationships between objects and task attributes to address affordance grasp detection. Furthermore, we have also constructed a semantic knowledge graph to encode the mappings of affordance grasping, enabling the designed Graph Attention Network (GAT) to generalize tool usage reasoning for novel tasks under the guidance of semantic information. Evaluations on the TaskGrasp dataset demonstrate that GATGrasp outperforms the established baselines and state-of-the-art methods, while experiments conducted on a real robot further verify the effectiveness of our method in performing taskaware affordance grasps on novel tools.

The integration of high-quality datasets, a generalized network model, and robust evaluation strategies sets a significant benchmark for advancing policy development in industrial bin-picking. This paper introduces the concept of region-aware grasping, a cutting-edge simulation to reality system designed to generate and evaluate 6D poses, empowering robots to grasp novel workpieces in stacked environments. The proposed system comprises two core components: the Sim2Real dataset, a large-scale synthetic point cloud dataset for grasp analysis, and Semantic-GraspNet, a policy framework that predicts full 6D grasp poses for stacked objects. By encoding and decoding point cloud data, Semantic-GraspNet innovatively transforms the pose prediction into a semantic categorization problem. Furthermore, we present a hybrid evaluation strategy that integrates pose assessment with mechanical grasp performance analysis, thereby enhancing both grasp success rates and sorting efficiency. To extend its capabilities, Semantic-GraspNet is combined with multi-modal large models, enabling accurate object-category-specific grasping in complex bin-picking scenarios. In real-world industrial applications, the system achieves a grasp completion rate of 91.3% in cluttered scenes and 89.2% in densely stacked environments, showcasing state-of-the-art performance in robotic picking and placing tasks.

Significant wave height (SWH) plays a critical role in marine operations, ship navigation, and climate prediction. The X-band radar is widely used for SWH inversion due to its short wavelength and high attenuation rate. Current inversion methods mainly fall into two categories: spatiotemporal-domain methods based on deep learning, and frequency-domain methods based on physical models. However, traditional deep learning methods typically emphasize spatial features while neglecting sequential information in radar image processing, which results in the loss of temporal features. Additionally, physics-based inversion methods in frequency-domain rely on manual feature design and cannot adaptively learn the nonlinear mapping between frequency feature and SWH, which hinders the improvement of inversion accuracy. To solve these problems, this paper proposes a bimodal cross-attention spatiotemporal-frequency fusion inversion network (BiCross-STFNet) to estimate SWH. First, the radar image sequences are simultaneously input into the spatiotemporal feature extraction module and frequency feature extraction module, where spatiotemporal and frequency features are extracted using the 3D spatiotemporal aware residual block. These features are then aligned and fused through the bimodal cross-attention fusion module, and finally, SWH is estimated. Experimental results demonstrate that BiCross-STFNet achieves a correlation coefficient of 0.964, a root mean square deviation (RMSD) as low as 0.04, and a mean absolute percentage error (MAPE) of 5.59%, outperforming existing methods.

This article presents a novel path planning approach for robotic manipulators operating in complex factory farming environments, where traditional methods struggle with strict obstacle avoidance constraints. The proposed method strikes a balance between minor permissible collisions and efficient obstacle avoidance. First, scene point clouds are downsampled using voxelization to generate a cost map. A greedy search is then employed to determine an initial obstacle-aware Cartesian path from this map. After postprocessing, the Cartesian path is converted into joint configuration trajectories using Ranged-IK, ensuring smooth, singularity-free transitions with controlled flexibility. The resulting validated trajectories are executed by the manipulator. Experiments were conducted on two robotic manipulators for pollination and harvesting tasks. The results indicate that the proposed method outperforms common alternatives, achieving higher operational efficiency, success rates, and adaptability, while permitting minor collisions.

The operation of traditional multi-object trackers on a moving autonomous aerial vehicle (AAV) faces many difficulties due to the irregular motion of AAV, the occlusion problem, and in particular arbitrarily oriented targets that are densely distributed with complex backgrounds. To solve these difficulties, this paper proposes a novel multi-object tracking framework, namely ArbiTrack, for a moving AAV to effectively detect and track arbitrarily oriented targets on the grounds. The proposed framework consists of an oriented object detection module to capture ground objects, a multi-scale context aggregation (MCA) module to improve the detection accuracy of small objects, and an adaptive motion switching (AMS) module to deal with the nonlinear complexity among AAV and ground objects. Historical information from multiple moments is used in this framework to learn the spatio-temporal characteristics so that the occlusion problem can be solved effectively. Experiments are conducted by using our OriDrone dataset and the public dataset UAVDT dataset. Results demonstrate that the proposed method achieves state-of-the-art tracking performance.

This paper presents a hierarchical LiDAR simulation framework to address the challenges of accurately simulating LiDAR data in autonomous driving scenarios. The framework utilizes a homology mapping approach to integrate LiDAR responses hierarchically at three levels: the instantaneous power response, environmental optical channel response, and target reflection response of LiDAR. This allows for the dynamic coupling of LiDAR geometric and physical models with varying environmental parameters. By integrating an array of interactions intrinsic to the LiDAR system and its external environment, the proposed model can provide high-fidelity LiDAR point cloud simulations. The effectiveness of the simulated point clouds has been validated through extensive experiments using actual LiDAR data and detection algorithms trained on existing datasets. The experimental results show that the proposed method has the potential to improve the realism of LiDAR simulations and the accumulation of challenging perception data.

Unmanned ground vehicles (UGVs) require effective perception and analysis of their surrounding terrain for safe operation. This paper presents a novel approach to their local elevation mapping and traversability analysis using sparse data from a single LiDAR sensor, which can generate a dense local traversability map in real-time. By preserving ground height information, our method can differentiate between vertical obstacles, suspended objects and other terrains in the elevation map. The modified Bayesian generalized kernel elevation inference is utilized to predict and fill in sparse elevation maps to achieve local dense terrain traversability mapping. The system uses GPU parallel processing to accelerate calculations, ensuring real-time perception and dynamic processing. The proposed system was tested in both structured and unstructured environments, and achieved better performances in map filling and handling of suspended and vertical objects compared to other existing approaches.

Depth estimation from monocular video plays a crucial role in scene perception. The significant drawback of supervised learning models is the need for vast amounts of manually labeled data (ground truth) for training. To overcome this limitation, unsupervised learning strategies without the requirement for ground truth have achieved extensive attention from researchers in the past few years. This paper presents a novel unsupervised framework for estimating single-view depth and predicting camera motion jointly. Stereo image sequences are used to train the model while monocular images are required for inference. The presented framework is composed of two CNNs (depth CNN and pose CNN) which are trained concurrently and tested independently. The objective function is constructed on the basis of the epipolar geometry constraints between stereo image sequences. To improve the accuracy of the model, a left-right consistency loss is added to the objective function. The use of stereo image sequences enables us to utilize both spatial information between stereo images and temporal photometric warp error from image sequences. Experimental results on the KITTI and Cityscapes datasets show that our model not only outperforms prior unsupervised approaches but also achieving better results comparable with several supervised methods. Moreover, we also train our model on the Euroc dataset which is captured in an indoor environment. Experiments in indoor and outdoor scenes are conducted to test the generalization capability of the model.

Convolutional neural networks (CNNs) for 3-D data analyses require a large size of memory and fast computation power, making real-time applications difficult. This article proposes a novel OctreeNet (a sparse 3-D CNN) to analyze the sparse 3-D laser scanning data gathered from outdoor environments. It uses a collection of shallow octrees for 3-D scene representation to reduce the memory footprint of 3-D-CNNs and performs point cloud classification on every single octree. Furthermore, the smallest non-trivial and non-overlapped kernel (SNNK) implements convolution directly on the octree structure to reduce dense 3-D convolutions to matrix operations at sparse locations. The proposed neural network implements a depth-first search algorithm for real-time predictions. A conditional random field model is utilized for learning global semantic relationships and refining point cloud classification results. Two public data sets (Semantic3D.net and Oakland) are selected to test the classification performance in outdoor scenes with different spatial sparsity. The experiments and benchmark test results show that the proposed approach can be effectively used in real-time 3-D laser data analyses. Note to Practitioners-This article was motivated by the limitations of existing deep learning technologies for analyzing 3-D laser scanning data. This technology enables robots to infer what the surroundings are, which is closely linked to semantic mapping and navigation tasks. Previous deep neural networks have seldom been used in robotic systems since they require a large amount of memory and fast computation power to apply dense 3-D operations. This article presents a sparse 3-D-Convolutional neural network (CNN) for real-time point cloud classification by exploiting the sparsity of 3-D data. This framework requires no GPUs. The practicality of the proposed method is verified on data sets gathered from different platforms and sensors. The proposed network can be adopted for other classification tasks with laser sensors.

Multiple images have been widely used for scene understanding and navigation of unmanned ground vehicles in long term operations. However, as the amount of visual data in multiple images is huge, the cumulative error in many cases becomes untenable. This paper proposes a novel method that can extract features from a large dataset of multiple images efficiently. Then the membership K-means clustering is used for high dimensional features, and the large dataset is divided into N subdatasets to train N conditional random field (CRF) models based on superpixel. A Softmax subdataset selector is used to decide which one of the N CRF models is chosen as the prediction model for labeling images. Furthermore, some experiments are conducted to evaluate the feasibility and performance of the proposed approach.

Conventional Chinese calligraphy robots often suffer from the limited sizes of predefined font databases, which prevent the robots from writing new characters. This paper presents a robotic handwriting system to address such limitations, which extracts Chinese characters from textbooks and uses a robot's manipulator to write the characters in a different style. The key technologies of the proposed approach include the following: 1) automatically decomposing Chinese characters into strokes using Harris corner detection technology and 2) matching the decomposed strokes to robotic writing trajectories learned from human gestures. Briefly, the system first decomposes a given Chinese character into a set of strokes and obtains the stroke trajectory writing ability by following the gestures performed by a human demonstrator. Then, it applies a stroke classification method that recognizes the decomposed strokes as robotic writing trajectories. Finally, the robot arm is driven to follow the trajectories and thus write the Chinese character. Seven common Chinese characters have been used in an experiment for system validation and evaluation. The experimental results demonstrate the power of the proposed system, given that the robot successfully wrote all the testing characters in the given Chinese calligraphic style.

A quadrotor unmanned aerial vehicle (UAV) should have the ability to perform real-time target tracking and path planning simultaneously even when the target enters unstructured scenes, such as groves or forests. To accomplish this task, a novel system framework is designed and proposed to accomplish simultaneous moving target tracking and path planning by a quadrotor UAV with an onboard embedded computer, vision sensors, and a two-dimensional laser scanner. A support vector machine-based target screening algorithm is deployed to select the correct target from multiple candidates detected by single shot multibox detector. Furthermore, a new tracker named TLD-KCF is presented in this paper, in which a conditional scale adaptive algorithm is adopted to improve the tracking performance for a quadrotor UAV in cluttered outdoor environments. According to distance and position estimation for a moving target, our quadrotor UAV can acquire a control point to guide its fight. To reduce the computational burden, a fast path planning algorithm is proposed based on elliptical tangent model. A series of experiments are conducted on our quadrotor UAV platform DJI M100. Experimental video and comparison results among four kinds of target tracking algorithms are given to show the validity and practicality of the proposed approach.

Currently cyber-security has attracted a lot of attention, in particular in wireless industrial control networks (WICNs). In this paper, the stability of wireless networked control systems (WNCSs) under deception attacks is studied with a token-based protocol applied to the data link layer (DLL) of WICNS. Since deception attacks cause the stability problem of WNCSs by changing the data transmitted over wireless network, it is important to detect deception attacks, discard the injected false data and compensate for the missing data (i.e., the discarded original data with the injected false data). The main contributions of this paper are: (1) With respect to the character of the token-based protocol, a switched system model is developed. Different from the traditional switched system where the number of subsystems is fixed, in our new model this number will be changed under deception attacks. (2) For this model, a new Kalman filter (KF) is developed for the purpose of attack detection and the missing data reconstruction. (3) For the given linear feedback WNCSs, when the noise level is below a threshold derived in this paper, the maximum allowable duration of deception attacks is obtained to maintain the exponential stability of the system. Finally, a numerical example based on a linearized model of an inverted pendulum is provided to demonstrate the proposed design.

This paper presents an indoor relocalization system using a dual-stream convolutional neural network (CNN) with both color images and depth images as the network inputs. Aiming at the pose regression problem, a deep neural network architecture for RGB-D images is introduced, a training method by stages for the dual-stream CNN is presented, different depth image encoding methods are discussed, and a novel encoding method is proposed. By introducing the range information into the network through a dual-stream architecture, we not only improved the relocalization accuracy by about 20% compared with the state-of-the-art deep learning method for pose regression, but also greatly enhanced the system robustness in challenging scenes such as large-scale, dynamic, fast movement, and night-time environments. To the best of our knowledge, this is the first work to solve the indoor relocalization problems based on deep CNNs with RGB-D camera. The method is first evaluated on the Microsoft 7-Scenes data set to show its advantage in accuracy compared with other CNNs. Large-scale indoor relocalization is further presented using our method. The experimental results show that 0.3 m in position and 4° in orientation accuracy could be obtained. Finally, this method is evaluated on challenging indoor data sets collected from motion capture system. The results show that the relocalization performance is hardly affected by dynamic objects, motion blur, or night-time environments.

This paper investigates the problem of multiclass and multiview 3-D object detection for service robots operating in a cluttered indoor environment. A novel 3-D object detection system using laser point clouds is proposed to deal with cluttered indoor scenes with a fewer and imbalanced training data. Raw 3-D point clouds are first transformed to 2-D bearing angle images to reduce the computational cost, and then jointly trained multiple object detectors are deployed to perform the multiclass and multiview 3-D object detection. The reclassification technique is utilized on each detected low confidence bounding box in the system to reduce false alarms in the detection. The RUS-SMOTEboost algorithm is used to train a group of independent binary classifiers with imbalanced training data. Dense histograms of oriented gradients and local binary pattern features are combined as a feature set for the reclassification task. Based on the dalian university of technology (DUT)-3-D data set taken from various office and household environments, experimental results show the validity and good performance of the proposed method.

Robotic dance is an important topic in the field of social robotics. Its research has a vital significance to both humans and robotics. This paper presents a review of the state of the art in robotic dance. Robotic dance is classified into four categories: cooperative human-robot dance, imitation of human dance motions, synchronization for music, and creation of robotic choreography. The research methods in each category are discussed. Future research areas are highlighted.

A potential threat to embedded systems is the execution of unknown or malicious software capable of triggering harmful system behavior, aimed at theft of sensitive data or causing damage to the system. Commercial off-the-shelf embedded devices, such as embedded medical equipment, are more vulnerable as these type of products cannot be amended conventionally or have limited resources to implement protection mechanisms. In this paper, we present a self-organizing map (SOM)-based approach to enhance embedded system security by detecting abnormal program behavior. The proposed method extracts features derived from processor's program counter and cycles per instruction, and then utilises the features to identify abnormal behavior using the SOM. Results achieved in our experiment show that the proposed method can identify unknown program behaviors not included in the training set with over 98.4% accuracy.

This paper presents a novel robotic fish, iSplash-I, with full-body coordination and high performance carangiform swimming motion. The proposed full-body length swimming motion coordinates anterior, mid-body and posterior displacements in an attempt to reduce the large kinematic errors in the existing free swimming robotic fish. It optimizes forces around the center of mass and initiates the starting moment of added mass upstream. A novel mechanical drive system was devised operating in the two swimming patterns. Experimental results show, that the proposed carangiform swimming motion approach has significantly outperformed the traditional posterior confined undulatory swimming pattern approach in terms of the speed measured in body lengths/ second, achieving a maximum velocity of 3.4BL/s and consistently generating a velocity of 2.8BL/s at 6.6Hz.

This paper studies three-dimensional multi-robot Cooperative Localization (CL) problem. Most of existing CL strategies adopt Extended Kalman Filter (EKF) or Maximum a Posteriori (MAP). In this paper, a novel approach based on Moving Horizon Estimation (MHE) is proposed. The main contribution of this paper is twofold: 1) MHE is integrated with EKF for three-dimensional CL using single mobile beacon, which can bound localization error, impose various constraints on states and noises, and make use of previous range measurements for current estimation. 2) A sufficient condition on observability of multi-robot CL is derived by using Fisher Information Matrix. Simulation is conducted to verify that the proposed MHE based CL algorithm outperforms EKF based method in terms of localization accuracy, and two scenarios where our algorithm is superior to EKF are discussed.

This paper presents a millimeter scale robotic fish, namely iSplash-MICRO, able to accurately generate the posterior undulatory pattern of the carangiform swimming mode, at intensively high frequencies. Furthermore an investigation into anterior stabilization was made in an attempt to reduce the large kinematic errors and optimize forces around the center of mass. Applying large scale dorsal and pelvic fins relative to body size enabled predictable optimization of the anterior and posterior displacements. During the field trials, the small fish with a length of 50mm has generated an equivalent average maximum velocity to real fish, measured in body lengths/ second (BL/s), greatly improving previous man-made systems, achieving a consistent free swimming speed of 10.4BL/s (0.52m/s) at 19Hz with a low energy consumption of 0.8 Watts.

This paper introduces a new robotic fish, iSplash-II, capable of outperforming real carangiform fish in terms of average maximum velocity (measured in body lengths/ second) and endurance, the duration that top speed is maintained. A new fabrication technique and mechanical drive system were developed, effectively transmitting large forces at high frequencies to obtain high-speed propulsion. The lateral and thrust forces were optimized around the center of mass, generating accurate kinematic displacements and greatly increasing the magnitude of added mass. The prototype, with a length of 32cm has significantly increased the linear swimming speed of robotic fish, achieving consistent untethered stabilized swimming speeds of 11.6BL/s (i.e. 3.7m/s), with a frequency of 20Hz.

This paper studies the underwater localization problem for a school of robotic fish, i.e., a kind of Autonomous Underwater Vehicles with limited size, power and payload. These robotic fish cannot be equipped with traditional underwater localization sensors that are big and heavy. The proposed localization system is performed by using a single surface mobile beacon which provides range measurement to bound the localization error. The main contribution of this paper lies in twofold: 1) Observability of single beacon based localization is first analyzed in the context of nonlinear discrete time system, deriving a sufficient condition on observability. 2) Moving Horizon Estimation is then integrated with Extended Kalman Filters for three-dimensional localization using single beacon, which can reduce the computational complexity, impose various constraints and make use of previous range measurements for current estimation. Extensive numerical simulations are conducted to verify the observability and high localization accuracy of the proposed underwater localization method.

The demographic trend is towards ageing in our society and the number of people with physical impairments and disabilities will increase dramatically in the future. It is necessary to deliver advanced healthcare and services to these people so that they can live independently and stay well at home throughout their lifespan. This paper presents a multi-robot architecture for ambient assisted living of the elderly and disabled, which is based on the robot operating system (ROS). A communication bridge is proposed for different means of human robot interaction, and ROS provides a framework for rapid system development with a reduced cost. Some experimental results are given in the paper to demonstrate the feasibility and performance of the proposed system.

The volume of prostate is one of the key indicators in the diagnosis and treatment of prostate cancer. This paper presents a novel method to calculate the volume of prostate in MRI images with high accuracy and in real time. In this approach, wavelet transform is used to decompose a MRI image into spatially oriented channels and then decompose each sub-image into 1D signal, by obtaining integral of sub-images. Bayesian decision theory is then used to analyze signals and detect the boundaries of prostate. Experimental results show that the proposed method can be implemented in real time and has acceptable accuracy.

The pivotal idea of recognizing human aggressive behaviors underlines how a taxonomer models such actions to perform recognition. In this paper, we investigate both the recognition and modeling of aggressive behaviors using kinematic (3-D) and electromyographic performance data. For this purpose, the Gaussian ground-plan projection area model has been assessed as an excellent evolutionary paradigm for the multiclass action and behavior recognition problem. In fact, it has shown superior classification accuracy with and without the use of ensemble models compared with the standard Gaussian (distance and area) models and other metrics of divergence, when dedicated groups of actions (behaviors) are being modeled. Genetic Programming is being employed to construct behavior-based taxonomers with a biomechanical primitive language. The modeling process revealed a representative subset of parameters (limbs, body segments, and marker coordinates) that are selected through the evolutionary process.

Providing a visual image of a hazardous substance such as nerve gas or nuclear radiation using multiple robotic agents could be very useful particularly when the substance is invisible. Such visual representation could show where the hazardous substance concentration is highest through the deployment of a higher density of robotic agents to that area enabling humans to avoid such areas. We present an algorithm that is capable of doing the aforementioned with very minimal cost when compared with other techniques such as Voronoi partition methods. Using a mathematical proof, we show that the algorithm would always converge to the distribution of a spatial quantity under investigation. The mathematical model of the bacterium as developed by Berg and Brown is used in this paper, and through simulations and physical experiments, we show that a controller based upon the model is capable of being used to visually represent an invisible spatial hazardous substance using simplistic agents with the future possibility of the same algorithm being used to track a rapidly changing spatiotemporal substance. We believe that the algorithm has this potential because of its low communication and computational needs.

The real-time adaptation between human and assistive devices can improve the quality of life for amputees, which, however, may be difficult to achieve since physical and mental states vary over time. This paper presents a co-adaptive human-machine interface (HMI) that is developed to control virtual forearm prosthesis over a long period of operation. Direct physical performance measures for the requested tasks are calculated. Bioelectric signals are recorded using one pair of electrodes placed on the frontal face region of a user to extract the mental (affective) measures (the entropy of the alpha band of the forehead electroencephalography signals) while performing the tasks. By developing an effective algorithm, the proposed HMI can adapt itself to the mental states of a user, thus improving its usability. The quantitative results from 16 users (including an amputee) show that the proposed HMI achieved better physical performance measures in comparison with the traditional (nonadaptive) interface ({\rm p\hbox{-}value}<0.001). Furthermore, there is a high correlation (correlation coefficient < 0.9, {\rm p\hbox{-}value} < .01) between the physical performance measures and self-report feedbacks based on the NASA TLX questionnaire. As a result, the proposed adaptive HMI outperformed a traditional HMI.

This paper presents a method of using Gaussian process regression to model spatial functions for mobile wireless sensor networks. A distributed Gaussian process regression (DGPR) approach is developed by using a sparse Gaussian process regression method and a compactly supported covariance function. The resultant formulation of the DGPR approach only requires neighbor-to-neighbor communication, which enables each sensor node within a network to produce the regression result independently. The collective motion control is implemented by using a locational optimization algorithm, which utilizes the information entropy from the DGPR result. The collective mobility of sensor networks plus the online learning capability of the DGPR approach also enables the mobile sensor network to adapt to spatiotemporal functions. Simulation results are provided to show the performance of the proposed approach in modeling stationary spatial functions and spatiotemporal functions.

In this paper, a survey is being conducted on the investigation of a four-class taxonomy related to security robots that appeared over the past three decades. The survey emphasizes on state-of-the-art mobile technologies that have been developed for crime-fighting robots, capable of crafting critical situations with confrontation strategies. Throughout this investigation, 60 projects are being examined with respect to faculties and sensor apparatus being used. A statistical analysis, which is carried on the historical developments of the most attractive frameworks, reveals the popularity of the four security robot categories and their chronological progress over the past 30 years. The categories being evaluated regard teleoperated, distributed, surveillance, and law-enforcement robot architectures. In the survey, an attempt is made to explain the importance of intelligent methodologies, and their emergent effects in security tasks. The major findings of this analysis illustrate the minor contribution of intelligent architectures in crime-fighting robots, and what constitutes an intelligent security robot.

To understand and interpret human motion is a very active research area nowadays because of its importance in sports sciences, health care, and video surveillance. However, classification of human motion patterns is still a challenging topic because of the variations in kinetics and kinematics of human movements. In this paper, we present a novel algorithm for automatic classification of motion trajectories of human upper limbs. The proposed scheme starts from transforming 3-D positions and rotations of the shoulder/elbow/wrist joints into 2-D trajectories. Discriminative features of these 2-D trajectories are, then, extracted using a probabilistic shape-context method. Afterward, these features are classified using a k-means clustering algorithm. Experimental results demonstrate the superiority of the proposed method over the state-of-the-art techniques.

In this paper, an investigation of evolvable probabilistic classifiers is conducted, along with a thorough comparison between a classical Gaussian distance model, and the induction of Gaussian-to-circle projection model. The newly introduced model refers to a distance fitness measure, based on the projection of Gaussian distributions with geometric circles. The projection architecture aims to model and classify physical aggressive behaviours, by using biomechanical primitives. The primitives are being used to model the dynamics of the aggressive activities, by evolving biomechanical classifiers, which can discriminate between three behaviours and six actions. Both evolutionary models have shown strong discrimination performances on recognising the individual actions of each behaviour. From the comparison, the proposed model outperformed the classical one with three ensemble programs.

Inspired by the simplicity of how nature solves its problems, we implement a bacteria controller on a physical platform that would enable the localisation and subsequent mapping of environmental pollution. We investigate the effects of each parameter in the controller on the localisation and exploration ability of the platform used. We also present how we can tune the controller for a given environmental condition depending on whether localisation or exploration is of a major priority. Some experimental results are presented to show the feasibility and performance of the proposed bacteria control.

In this paper, we investigate a moving target tracking problem with mobile sensor networks. The moving target is assumed to be an intelligent agent, which is ‘smart’ enough to escape from the detection. We formulate this target estimation problem as a zero-sum game in this paper and use a so-called minimax filter to estimate the target position. The minimax filter is a robust filter that minimizes the estimation error by considering the worst case noise. Furthermore we develop a distributed version of the minimax filter for multiple sensor nodes. The distributed computation is implemented via a consensus filter. Finally, the mobile sensor nodes need to control their motions to move towards the estimated target position and avoid collisions with neighbors. A flocking algorithm is developed for this purpose. The simulation results show that the target tracking algorithm proposed in this paper provides a satisfactory result.

This paper presents a distributed algorithm for mobile sensor networks to monitor the environment. With this algorithm, multiple mobile sensor nodes can collectively sample the environmental field and recover the environmental field function via machine learning approaches. The mobile sensor nodes are able to self-organise so that the distribution of mobile sensor nodes matches to the estimated environmental field function. In this way, it is possible to make the next-step sampling more accurate and efficient. The machine learning approach used for function regression is support vector regression (SV R) algorithm. A distributed SV R learning algorithm is used for on-line learning. The self-organised algorithm used for deployment is based on locational optimisation techniques. In particular, Lloyd's algorithm for optimising centroidal Voronoi tessellations (CV T) is used to spread mobile sensor nodes over the monitored environment. The environmental field function is simulated in static and dynamic settings and the demonstration on the simulated environments shows the proposed algorithm is effective.

A repertory of nine biomechanical aggressive activities is investigated in this paper, in our effort to instigate a new paradigm at aggregating descriptive mathematical models with evolutionary, symbolic program representations. Such representations are based on shared biomechanical primitives inspired from kinematics, dynamics, and energetics. Our intension is twofold, initially to study the nature of aggressive biomechanical models and then to classify their physical activities by evolving expression-trees with biomechanical synthesis. The methodology targets on evolving expression programs using the Gaussian Ground-plan Projection Area model, to discriminate among three aggressive behaviours and recognise the individual actions involved. For the n-class problem, three programs have been evolved, each for an aggressive behaviour such as the arm-Launch, the legLaunch, and the bodyLaunch behaviour, so that to be able to examine separately the evolvable characteristics induced. The proposed approach has evidently shown strong classification and discrimination performances.

One of fundamental issues for service robots is human-robot interaction. In order to perform such a task and provide the desired services, these robots need to detect and track people in the surroundings. In this paper, we propose a solution for human tracking with a mobile robot that implements multisensor data fusion techniques. The system utilizes a new algorithm for laser-based leg detection using the onboard laser range finder (LRF). The approach is based on the recognition of typical leg patterns extracted from laser scans, which are shown to also be very discriminative in cluttered environments. These patterns can be used to localize both static and walking persons, even when the robot moves. Furthermore, faces are detected using the robot's camera, and the information is fused to the legs' position using a sequential implementation of unscented Kalman filter. The proposed solution is feasible for service robots with a similar device configuration and has been successfully implemented on two different mobile platforms. Several experiments illustrate the effectiveness of our approach, showing that robust human tracking can be performed within complex indoor environments.

Two different classifier representations based on dynamic Artificial Neural Networks (ANNs) and Genetic Programming (GP) are being compared on a human action recognition task by an ubiquitous mobile robot. The classification methodologies used, process time series generated by an indoor ubiquitous 3D tracker which generates spatial points based on 23 reflectable markers attached on a human body. This investigation focuses mainly on class discrimination of normal and aggressive action recognition performed by an architecture which implements an interconnection between an ubiquitous 3D sensory tracker system and a mobile robot to perceive, process, and classify physical human actions. The 3D tracker and the robot are used as a perception-to-action architecture to process physical activities generated by human subjects. Both classifiers process the activity time series to eventually generate surveillance assessment reports by generating evaluation statistics indicating the classification accuracy of the actions recognized.

This paper proposes a myoelectric-based virtual joystick to manipulate an electric powered wheelchair for people with severe motor disabilities. It is built on pattern recognition-based myoelectric control that discriminates muscular activities using forearmpsilas surface myoelectric signals. The core of the system is a support vector machine based classifier that classifies signal time domain features. Online training scheme is applied to cope with gradual changes in myoelectric signal patterns. Two indexes, contineousness and entropy, were used to update training data set in real time operation. The results confirm that the proposed myoelectric-based joystick is a reliable alternative for traditional joysticks and its performance is fairly acceptable.

Flocking algorithms essentially consist of three components: alignment, cohesion, and separation. To track a desired trajectory, the flock center should move along the desired trajectory, and thus, the fourth component, navigation, is necessary. The alignment, cohesion, and navigation components are well implemented through consensus protocols and tracking controls, while the separation component is designed through heuristic-based approaches. This paper proposes a fuzzy logic solution to the separation component. The TS rules and Gaussian membership functions are used in fuzzy logic. For fixed network flocking, a standard stability proof by using LaSalle's invariance principle is provided. For dynamic network flocking, a Filipov solution definition is given for nonsmooth dynamics. Then, a LaSalle's invariance principle for nonsmooth dynamics is used to prove the stability. A group of mobile robots with double integrator dynamics is simulated for the flocking algorithms in a 2-D environment.

This paper presents the design and implementation of a coordination architecture for quadruped walking robots to learn and execute soccer-playing behaviors. A typical hybrid architecture combing reactive behaviors with deliberative reasoning is developed. The reactive behaviors directly map spatial information extracted from sensors into actions. The deliberative reasoning represents temporal constraints of a robot's strategy in terms of finite state machines. In order to achieve real-time and robust control performance in reactive behaviors, fuzzy logic controllers (FLCs) are used to encode the behaviors, and a two-stage learning scheme is adopted to make these FLCs adaptive to complex situations. The experimental results are provided to show the suitability of the architecture and effectiveness of the proposed learning scheme.

To realise fish-like swim patterns by a robotic system poses tremendous challenges. This requires fully understanding of fish biomechanics and the way to mimic it. This paper presents our research toward the sensor-based control of autonomous robotic fish that can swim in a 3D unstructured environment, based on observations of fish swim behaviours. Our robotic fish has a tail with three or four degrees of freedom (DOF) and is controlled by 4 onboard computers (a powerful Gumstix Linux PC and 3 PICs) and over 10 embedded sensors. Both simulated and the real fish experiments are conducted to show the feasibility and performance of the proposed approach

This paper presents a hybrid control architecture for autonomous robotic fishes which are able to swim and navigate in unknown or dynamically changing environments. It has a three-layer configuration: cognitive layer, behaviour layer and swim pattern layer. The state-based planning in the cognitive layer provides a good foundation for potential adaptation through machine learning methods such as reinforcement learning (RL). The behaviour layer and the swim pattern layer are specially designed to match the needs for the real-time control of our robotic fish. To test the feasibility and performance of the proposed architecture, the experiment of "tank border exploration" is conducted with Q-learning

This paper presents a novel Fuzzy Sarsa(λ) Learning (FSλL) approach applied to a strategic route leaning task of a mobile robot. FSlambdaL is a hybrid architecture that combines reinforcement learning and fuzzy logic control. The Sarsa(λ) learning algorithm is used to tune the rule-base of a fuzzy Logic controller which has been tested in a route learning task. The robot explores its environment using its fixed experience provided by a discretized fuzzy logic controller, and then learns optimal policies to achieve goals in less time and less error.

This paper presents a new approach for image segmentation by combining the classical gradient vector flow (GVF) algorithm with mean shift. Due to the dependence on the gradient vectors of an edge map, the classical GVF is sensitive to the shape irregularities, and hence the snake cannot be ideally located on the concave boundaries. We propose an improved representation of the internal energy force by reducing the Euclidean distance between the guessed centroid and the estimated one of the snake. Experimental work shows the performance of this approach in different tests.

In nature, fish has astonishing swimming ability after thousands years of evolution. To realise fish-like swimming behaviours by a robotic system poses tremendous challenges, especially for the C-shape turning (CST). This requires fully understanding of fish biomechanics and the way to mimic it. Based on observations of fish swimming, this paper presents a new kinematics model to mimic the CST behaviour in a robotic fish with a 4-DOF (degrees of freedom) tail. The simulated and the real experiments are conducted to show the theoretic feasibility. Both behaviour analyses and hydrodynamics features between the robotic fish and the real fish are presented to show the performance.

Map building through cooperative localisation (co-location) using circular geometric targets and a SICK laser range scanner is investigated. The tenet of co-location is circle detection in laser range data. Two methods for circle detection, a Range Weighted Circular Hough Transform (RWCHT) and a novel squared-residual voting strategy are compared and their performance assessed. The custom squared-residual voting strategy outperforms the RWCHT in all respects and is subsequently used for localisation and map building. The results include robust continuous localisation at speeds of 0.2m/s with 98% of scan frames used and an error of less than 0.03m. This localisation accuracy helps build maps of 96% quality and occupancy grids of cluttered environments despite the presence of distractors.

This paper presents a novel mechatronics design for a 3D swimming robotic fish, namely MT1 (Mechanical Tail) robotic fish. It has a novel tail structure which uses only one motor to generate fish-like swimming motion using C-bends tail shapes. This design enables MT1 to become the first small size robotic fish (<0.5m in length) and be able to dive over 3 meters deep in water. An effective control method with only 5 parameters is proposed to control its 3D swimming behaviours. Experimental results are presented to show the feasibility and good performance of the proposed control algorithms.

This paper addresses the control of a leader-follower formation where the leader robot has its own target and the follower robots are constrained by the specified formation tasks. The dynamics of the leader robot with nonholonomic constraint is explicitly integrated into the formation system to yield a centralized coordinating controller. As a result there is no need to assume the motion of the leader separatively when we develop cooperative formation controllers for coordinating the robots. The feedback linearization is used to deal with the nonlinear formation control of a team of autonomous mobile robots with nonholonomic constraints. Although the nonlinear formation system under consideration can be exactly linearized by taking advantage of dynamic feedback linearization, there exists structural singularity which may pose serious problems in practice. To solve this singular problem a new formation model for controlling the leader-follower formation in a cooperative manner is developed. This new formation model can be extended to studying other control and learning issues in multi-robot systems for both cooperation and noncooperation. The internal dynamics is derived and proven to be globally stable under the stable linear controller obtained via the partially linearized dynamics. To demonstrate the performance of the developed formation controller, simulation results are provided.

This article presents a hybrid learning architecture for fuzzy control of quadruped walking robots in the RoboCup domain. It combines reactive behaviors with deliberative reasoning to achieve complex goals in uncertain and dynamic environments. To achieve real-time and robust control performance, fuzzy logic controllers (FLCs) are used to encode the behaviors and a two-stage learning scheme is adopted to make these FLCs be adaptive to complex situations. The first stage is called structure learning, in which the rule base of an FLC is generated by a Q-learning scheme. The second stage is called parameter learning, in which the parameters of membership functions in input fuzzy sets are learned by using a real value genetic algorithm. The experimental results are provided to show the suitability of the architecture and effectiveness of the proposed learning scheme. © 2005 Wiley Periodicals, Inc. Int J Int Syst 20: 131–152, 2005.

This paper presents a receding horizon (RH) controller used for regulating a nonholonomic mobile robot. The RH control stability is guaranteed by adding a terminal-state penalty to the cost function and a terminal-state region to optimization constraints. A suboptimal solution to the optimization problem is sufficient to achieve stability. A new terminal-state penalty and its corresponding terminal-state constraints are found. Implementation and simulation results are provided to verify the proposed control strategy.

This work presents a learning approach to fuzzy classifier systems. Q-learning algorithm is employed to implement credit assignment of the learning. GA operators are used as an action selection mechanism of the learning. The learning approaches can be viewed as a fuzzy learning classifier system or a Q-learning algorithm that adopts fuzzy logic to generalise Q-learning results. Rule accuracies are treated as rule fitness values. The learning algorithm is applied to a control robot behaviour.

Behaviour co-ordination is one of the major problems in behaviour-based robotics. This paper presents a teaching method for mobile robots to learn behaviour coordination. In this method, the sensory information is abstracted into a limited number of feature states that correspond to physical events in the interactive process between a robot and its environment. The continuous motor actions are abstracted into a limited number of behaviours. The goal of the behaviour co-ordination is to map the feature states into the behaviours in the light of environment rewards. The teaching process consists of an imitation stage and an autonomous learning stage. Both stages employ Q-learning algorithms to implement the mapping. The imitation stage serves as a preliminary stage for the teaching method. The learning result is used to bootstrap the autonomous learning stage. Experiments are conducted in the domain of soccer playing by Sony legged robots. Experiment results show that the robots can acquire behaviour coordination ability.

We presented a new learning approach to the application of service robots, which is based on learning by imitation. Service robots need to increase their set of actions, which would lead to the ability of adapting their behaviours. In contrast with traditional learning approaches learning by imitation presents considerable advantages; equip robots with the abilities to be efficient in applications requiring human interaction. The paper offers our experiences with the first stage of our approach. Experimental results show the feasibility of such an approach.

This paper presents a 3D simulator used for studying the motion control and autonomous navigation of a robotic fish. The simplified kinematics and hydrodynamics models are created for the simulator, including many other object models such as water, obstacles, sonar sensors and a swimming pool. The experimental results show that the use of this simulator is a realistic and convenient way to develop autonomous navigation algorithms for robotic fishes.

In the near future, container ports will no longer be able to expand into the surrounding land and will thus be unable to meet the storage requirements due to the boom in world trade. A solution to this problem is to increase the container throughput of the port by reducing the amount of time necessary to load and unload a ship. This paper presents a distributed agent architecture to achieve this task. Under such architecture, an intelligent planning algorithm is continuously optimised by the dynamic and co-operative rescheduling of yard resources such as quay cranes and container vehicles.

The complexity of most modem systems prohibits a hand-coded approach to decision making. In addition, many problems have continuous or large discrete state spaces; some have large or continuous action spaces. The problem of learning in large spaces is tackled through generalisation techniques, which allow compact representation of learned information and transfer of knowledge between similar states and actions. In this paper Kanerva. coding and reinforcement learning are combined to produce the KaBaGe-RL decision-making module. The purpose of KaBaGe-RL is twofold. Firstly, Kanerva coding is used as a generalisation method to produce a feature vector from the raw sensory input. Secondly, the reinforcement learning uses this feature vector in order to learn an optimal policy. The efficiency of KaBaGe-RL is tested using the "3 versus 2 possession football" challenge, a subproblem of the RoboCup domain. The results demonstrate that the learning approach outperforms a number of benchmark policies including a hand-coded one.

This paper presents a predictive control scheme for mobile robots that possess complexity, non-linearity and uncertainty. A multi-layer back-propagation neural network is employed as a model for nonlinear dynamics of the robot. The control variables are produced by optimizing the performance index on-line using the steepest gradient descent algorithm. The neural network is constructed by the wavelet orthogonal decomposition to form a wavelet neural network that can overcome the problems caused by local minima of optimization. The wavelet network is also helpful to determine the number of the hidden nodes and the initial value of weights. The sparse train data in our path tracking case can reduce the effect of the "curse of dimensionality" on the network size in high dimensional function learning caused by the orthogonal wavelet base function.

Developers of AI software are normally faced with design challenges involving robustness, efficiency, and extensibility. Most of these challenges at a higher level are independent of the application-specific requirements. Although design patterns have been successfully adopted to tackle these issues, they are rarely documented. Consequently this knowledge remains hidden in the minds of developers or buried within complex system source code. The primary contribution of the paper is to describe an abstract design methodology that can be applied in many (single or) multi-agent systems. The paper mainly illustrates how design patterns can ease the development and increase the efficiency of such systems. As an example, the Essex Wizards multi-agent system is presented which won the third prize in the RoboCup'99 simulator league competition.

The complexity of most multi-agent systems prohibits a hand-coded approach to decision-making. In addition to that a complex, dynamic, adversarial environment like the one of a football game makes decision-making and cooperation even more difficult. This paper addresses these problems by using machine learning techniques and agent technology. By gathering useful experience from earlier stages, an agent can significantly improve performance. The method used requires no previous knowledge regarding the environment. Since cooperation in adversarial domains is a very challenging task, the proposed learning algorithm assigns each agent a role to play to achieve a certain goal. By distributing the responsibilities among the agents and linking their goals, an efficient way of cooperation emerges.

Modern manufacturing requires flexible and autonomous systems for materials handling and transportation. For such applications, a succession of implemented mobile robots has been developed over the past seven years. In this paper, the questions of software and hardware architecture are discussed in light of experience gained so far. Also discussed are mobile robot projects that seem realistic over the next five to ten years.

We propose a probabilistic approach to the problem of global path planning with uncertainty for mobile robots in a dynamic manufacturing environment. To model the changing environment, we use a topological graph weighted by scalar cost functions. The cost functions consist of two elements: a deterministic cost for the known part of the robot's environment, and an uncertainty cost for the unknown part of the environment. Statistical models are built to quantify the unknown part of the environment, forming uncertainty costs for handling unexpected events. These uncertainty costs are dynamically updated by available sensor data when the mobile robot moves around. An optimal path (suboptimal in practice) is then found from the weighted topological graph using dynamic programming.

A modular architecture used for intelligent control of mobile robots has been developed at Oxford University over the last few years. This architecture takes the form of multiple sensing and control layers, based on "locally intelligent control agents" (LICAs). Central to such a design is the concept of "communication sequential processes" (CSP). This paper describes briefly the LICA-based control architecture for an Oxford mobile robot. Two typical operations of sonar and active vision systems have been given to demonstrate its flexibility and efficiency. Such a modular and unified system has been licensed commercially and allows for the integration of all kinds of sensors and actuators to meet future needs.

Investigates how a car-like mobile robot handles unexpected static obstacles while following an optimal path planned by the global path planner. To find an optimal solution of the problem, the obstacle avoidance problem is formulated as a decision theoretic approach. The optimal decision rule we seek is to minimize the Bayes risk by trading off between deliberative maneuver and the alternatives. Real-time implementation is emphasized in order to provide a framework for real-world applications.

Describes a decision theoretic approach to real-time obstacle avoidance and path planning for a mobile robot. The mobile robot navigates in a semi-structured environment in which unexpected obstacles may appear at random locations. Twelve sonar sensors are currently used to report the presence and location of the obstacles. To handle the uncertainty of an obstacle's appearance, the authors adopt a Bayesian approach by assuming a prior distribution for the presence of unknown obstacles. The distribution is changed dynamically according to the information accumulated by sensors. When searching for an optimal path using dynamic programming, the authors take the probability into account in making a decision. Based on prior information and sensor data, they show that the proposed method allows the mobile robot to avoid unexpected obstacles and finds an optimal path to the goal in real time.>

The design and partial implementation of a real-time architecture for a mobile robot, aimed particularly towards a vehicle developed for factory automation, is described. The authors develop a layered design to equip the robot with a number of behavioral competences. They examine sensing and a potential field algorithm especially to achieve modification of behavior at a speed close to the robot's operational speed. It is shown how the layered architecture interfaces to the original onboard architecture, which provided sophisticated localization but no ability to deal with environmental exceptions.>

A fully decentralized architecture is presented for data fusion problems. This architecture takes the form of a network of sensor nodes, each with its own processing facility, which together do not require any central processor or any central communication facility. In this architecture, computation is performed locally and communication occurs between any two nodes. Such an architecture has many desirable properties, including robustness to sensors failure and flexibility to the addition or loss of one or more sensors. This architecture is appropriate for the class of extended Kalman filter (EKF)-based geometric data fusion problems. The starting point for this architecture is an algorithm which allows the complete decentralization of the multisensor EKF equations among a number of sensing nodes. This algorithm is described, and it is shown how it can be applied to a number of different data-fusion problems. An application of this algorithm to the problem of multicamera, real-time tracking of objects and people moving through a room is described.>