Showing papers presented at "International Conference on Control, Automation, Robotics and Vision in 2018"

Map Comparison of Lidar-based 2D SLAM Algorithms Using Precise Ground Truth

Abstract: This paper presents a comparative analysis of three most common ROS-based 2D Simultaneous Localization and Mapping (SLAM) libraries: Google Cartographer, Gmapping and Hector SLAM, using a metrics of average distance to the nearest neighbor (ADNN). Each library was applied to construct a map using data from 2D lidar that was placed on an autonomous mobile robot. All the approaches have been evaluated and compared in terms of inaccuracy constructed maps against the precise ground truth presented by FARO laser tracker in static indoor environment.

Dimensionality Reduction for Machine Learning Based IoT Botnet Detection

Abstract: The rapid development of the internet of things caused severe security problems such as the cyber attacks launched by extremely huge botnets comprised of IoT devices. The detection of these devices is essential for protecting the networks. Recently, some of the studies have demonstrated the high accuracy of machine learning methods, including deep learning, in detecting IoT botnets. However, the minimizing of the required features for classification is highly needed for overcoming scalability and computation resource problems in IoT environments. Having results which can be readily interpretable by cyber security analysts and producing signatures for the contemporary intrusion detection or network monitoring systems are other significant factors in this area in which quick and widespread security adaption is highly required. In this study, we applied feature selection to minimize the number of features in detecting the IoT bots. It is shown that fewer features can achieve very high accuracy rates and afford interpretable results with a multi-class classifier based on a shallow method, decision tree.

RCCNet: An Efficient Convolutional Neural Network for Histological Routine Colon Cancer Nuclei Classification

Abstract: Efficient and precise classification of histological cell nuclei is of utmost importance due to its potential applications in the field of medical image analysis. It would facilitate the medical practitioners to better understand and explore various factors for cancer treatment. The classification of histological cell nuclei is a challenging task due to the cellular heterogeneity. This paper proposes an efficient Convolutional Neural Network (CNN) based architecture for classification of histological routine colon cancer nuclei named as RCCNet. The main objective of this network is to keep the CNN model as simple as possible. The proposed RCCNet model consists of 1, 512, 868 learnable parameters which are significantly less compared to the popular CNN models such as AlexNet, CIFAR-VGG, GoogLeNet, and WRN. The experiments are conducted over publicly available routine colon cancer histological dataset “CRCHistoPhenotypes”. The results of the proposed RCCNet model are compared with five state-of-the-art CNN models in terms of the accuracy, weighted average F1 score and training time. The proposed method has achieved a classification accuracy of 80.61% and 0.7887 weighted average F1 score. The proposed RCCNet is more efficient and generalized in terms of the training time and data over-fitting, respectively.

An Automated Irrigation System for Smart Agriculture Using the Internet of Things

Abstract: Water is a vital and scarce resource in agriculture and its optimal management is emerging as a key challenge. This paper presents an automated irrigation system to reduce water utilization in agriculture by combining the Internet of Things (IoT), cloud computing and optimization tools. The automated irrigation system deploys low cost sensors to sense variables of interest such as soil moisture, pH, soil type, and weather conditions. The data is stored in Thingspeak cloud service for monitoring and data-storage. The field data is transmitted to the cloud using Wi-Fi modem and using GSM cellular networks. Then an optimization model is used to compute the optimal irrigation rate which are automated using a solenoid valve controlled using an ARM controller (WEMOS D1). The variables of interest are stored in the cloud and offered as a service to the farmers. The proposed approach is demonstrated on a pilot scale agricultural facility and our results demonstrate the reduction in water utilization, increase in data-availability, and visualization.

Day-Ahead Price Forecasting for Electricity Market using Long-Short Term Memory Recurrent Neural Network

Abstract: In this paper, an efficient method for the day-ahead electricity price forecasting (EPF) is proposed based on a long-short term memory (LSTM) recurrent neural network model. LSTM network has been widely used in various applications such as natural language processing and time series analysis. It is capable of learning features and long term dependencies of the historical information on the current predictions for sequential data. We propose to use LSTM model to forecast the day-ahead electricity price for Australian market at Victoria (VIC) region and Singapore market. Instead of using only historical prices as inputs to the model, we also consider exogenous variables, such as holidays, day of the week, hour of the day, weather conditions, oil prices and historical price/demand, etc. The output is the electricity price for the next hour. The future 24 hours of prices are forecasted in a recursive manner. The mean absolute percentage error (MAPE) of four weeks for each season in VIC and Singapore markets are examined. The effectiveness of the proposed method is verified using real market data from both markets. The result shows that the LSTM network outperforms four popular forecasting methods and provides up to 47.3% improvement in the average daily MAPE for the VIC market.

A Two-step Method for Extrinsic Calibration between a Sparse 3D LiDAR and a Thermal Camera

Abstract: To obtain the 6 DOF extrinsic parameters (rotation and translation matrix) between a 3D ranging sensor and a thermal camera, previous methods require a high-resolution 3D ranging sensor to reliably detect features. Although sparse 3D LiDARs are widely used on autonomous robots, to the best of our knowledge, the extrinsic calibration between a sparse 3D LiDAR (particularly Velodyne VLP-16) and a thermal camera has not been considered in the literature. In this paper, we present a two-step method to address the problem, where a monocular visual camera is used to assist the process. The proposed method decomposes the problem into two steps: extrinsic calibration between a sparse 3D LiDAR and a visual camera; extrinsic calibration between a visual camera and a thermal camera. Experiments are conducted to demonstrate the effectiveness of the proposed two-step method.

Bi-Manual Articulated Robot Teleoperation using an External RGB-D Range Sensor

Abstract: In this paper, we present an implementation of a bi-manual teleoperation system, controlled by a human through three-dimensional (3D) skeleton extraction. The input data is given from a cheap RGB-D range sensor, such as the ASUS Xtion PRO. To achieve this, we have implemented a 3D version of the impressive OpenPose package, which was recently developed. The first stage of our method contains the execution of the OpenPose Convolutional Neural Network (CNN), using a sequence of RGB images as input. The extracted human skeleton pose localisation in two-dimensions (2D) is followed by the mapping of the extracted joint location estimations into their 3D pose in the camera frame. The output of this process is then used as input to drive the end-pose of the robotic hands relative to the human hand movements, through a whole-body inverse kinematics process in the Cartesian space. Finally, we implement the method as a ROS wrapper package and we test it on the centaur-like CENTAURO robot. Our demonstrated task is of a box and lever manipulation in real-time, as a result of a human task demonstration.

An improved Frontier-Based Approach for Autonomous Exploration

Abstract: A new approach for autonomous exploration in an unknown scenario based on the concept of frontiers is proposed in this paper. Exploration frontiers introduced by [4], are the regions on the boundary between open space and unexplored space. A mobile robot is able to construct its map by adding new space and moving to unvisited frontiers until the entire environment has been explored. However, the original frontier strategy, suffering from local minima, only considers distance and size of unknown spaces, resulting in low exploration efficiency in complex environments. By making use of the robot heading information using wheel odometry and coarse graph representation of the environment, the modified exploration method is able to balance the mapping coverage and time expenditure to a greater extent. The proposed method is experimentally verified on a mobile platform, exploring a real-world office environment cluttered with a variety of obstacles.

Real-Time Robot Vision on Low-Performance Computing Hardware

Abstract: Small robots have numerous interesting applications in domains like industry, education, scientific research, and services. For most applications vision is important, however, the limitations of the computing hardware make this a challenging task. In this paper, we address the problem of real-time object recognition and propose the Fast Regions of Interest Search (FROIS) algorithm to quickly find the ROIs of the objects in small robots with low-performance hardware. Subsequently, we use two methods to analyze the ROIs. First, we develop a Convolutional Neural Network on a desktop and deploy it onto the low-performance hardware for object recognition. Second, we adopt the Histogram of Oriented Gradients descriptor and linear Support Vector Machines classifier and optimize the HOG component for faster speed. The experimental results show that the methods work well on our small robots with Raspberry Pi 3 embedded 1.2 GHz ARM CPUs to recognize the objects. Furthermore, we obtain valuable insights about the trade-offs between speed and accuracy.

Improving UWB Based Indoor Positioning in Industrial Environments Through Machine Learning

Abstract: The detection and mitigation of Non-Line-of-Sight (NLOS) signals are crucial for achieving the full potential of UWB-based indoor positioning. In dense multipath industrial environments, it was seen that using the power characteristics of the received signal to identify NLOS conditions is effective when tracking stationary objects but is insufficient for mobile object tracking. Hence, machine learning classifiers utilizing Multi-Layer Perceptron (MLP) and Boosted Decision Trees (BDT) were developed to improve NLOS detection. Through experimental results from tests in a factory scenario, it is shown that BDT yields a higher accuracy of 87% as compared to the 79% obtained by the received power based method.

A Multi-Robot Task Allocation Method Based on Multi-Objective Optimization

Abstract: Time consumption and energy consumption are essential indicators for evaluating the effectiveness of task completion in multi-robot systems. On the basis of considering these two indicators, a Multi-Robot task allocation method based on multi-objective (time utility and energy utility) optimization (MOO-MRTA) is proposed. This method gives the definition and construction of the robotic energy utility function. It also establishes a model of task allocation based on multi-objective and discusses the problem of solving the model and so on. The RoboCup Rescue Simulation experiment shows that this method has the advantages of strong searching capacity, fast convergence rate, and that it can quickly find the Pareto optimal task allocation scheme and help the rescue team get an ideal result.

Vehicle Detection and Width Estimation in Rain by Fusing Radar and Vision

Abstract: While much effort has been devoted to deep learning object detection, relatively limited attention has been paid to object detection in bad weather, e.g. rain, snow or haze. In heavy rain, the raindrop on the front windshield can make it difficult to detect object from an in-car camera. The conventional way to cope with this has been to use radar as the main detection sensor. However, radar is highly susceptible to false positives. Furthermore, many entry level radar sensors only return the centroid of each detected object, rather than its size and extent. In addition, due to lack of texture input, radar cannot discriminate a vehicle from a non-vehicle object, e.g. roadside pole. This motivates us to detect vehicle by fusing radar and vision. In this paper, we first calibrate the radar and camera with respect to the ground plane. The radar detections are then projected to the camera image for target width estimation. Empirical evaluation on a large database shows that there is a natural synergy in both sensors, as the image based estimation is found to be greatly facilitated by the accuracy of the radar detection.

Short Text Classification with A Convolutional Neural Networks Based Method

Abstract: The traditional machine learning algorithms are easily affected by datasets in short text classification tasks, so they have weak generalization ability when confronted with new situations. This paper presents a new method SVMCNN by combining Convolutional Neural Networks and Support Vector Machine. Training the SVMCNN model with labeled datasets, and using the collected Twitter data for classification test. The results show that the SVMCNN, especially pre-trained SVMCNN has good performance in short text classification, which gets the high Precision rate, Recall rate and F1-measure.

A Comparative Study on Camera-Radar Calibration Methods

Abstract: Camera-radar fusion has been applied in obstacle detection or moving object tracking for autonomous vehicles and advanced driver assistance systems. When utilizing multiple sensors, their calibration is not only essential but also important because it gives great impacts on subsequent procedures. Nonetheless, camera-radar calibration methods have not been compared in the literature qualitatively or quantitatively. In this paper, we compare three types of the calibration methods presented in previous studies on fusing camera and radar in terms of calibration accuracy. Especially, the comparison is conducted in the situation of varying the number of radar-image data pairs used in their calibration. Experimental results show that one type of the methods is not appropriated to the camera-radar calibration, and the methods belonging to the other types provide quite similar accuracy.

Recurrent Deterministic Policy Gradient Method for Bipedal Locomotion on Rough Terrain Challenge

Abstract: This paper presents a deep learning framework that is capable of solving partially observable locomotion tasks based on our novel interpretation of Recurrent Deterministic Policy Gradient (RDPG). We study on bias of sampled error measure and its variance induced by the partial observability of environment and subtrajectory sampling, respectively. Three major improvements are introduced in our RDPG based learning framework: tail-step bootstrap of temporal difference, initialisation of hidden state using past subtrajectory, truncation of temporal backpropagation, and injection of external experiences learned by other agents. The proposed learning framework was implemented to solve the Bipedal-Walker challenge in OpenAI's gym simulation environment where only partial state information is available. Our simulation study shows that the autonomous behaviors generated by the RDPG agent are highly adaptive to a variety of obstacles and enables the agent to effectively traverse rugged terrains for long distance with higher success rate than leading contenders.

Lifelong Multi-Agent Path Finding in A Dynamic Environment

Abstract: In tradition, the problem of Multi-Agent Path Finding is to find paths for the agents without conflicts, and each agent execute one-shot task, a travel from a start position to its destination. However, making just one planning for the agents may not satisfy the requirement in dynamic environments such as logistics sorting center, where the paths of the agents may constantly need to be adjusted according to the incoming tasks. The challenging issue is to dynamically adjust the already planned paths while make planning for the agents ready to execute new incoming tasks. In this paper, we formulate it into Dynamic Multi-Agent Path Finding (DMAPF) problem, the goal of which is to minimize the cumulative cost of paths. To solve this problem, we propose an algorithm called Lifelong Planning Conflict-Based Search (LPCBS), which can efficiently and optimally make planning for the new incoming tasks while adjusting the already planned paths. Experiment results show that the LPCBS performs much better than the existing works in each planning.

Efficient Trajectory Optimization for Robot Motion Planning

Abstract: Motion planning for multi-jointed robots is challenging. Due to the inherent complexity of the problem, most existing works decompose motion planning as easier subproblems. However, because of the inconsistent performance metrics, only sub-optimal solution can be found by decomposition based approaches. This paper presents an optimal control based approach to address the path planning and trajectory planning subproblems simultaneously. Unlike similar works which either ignore robot dynamics or require long computation time, an efficient numerical method for trajectory optimization is presented in this paper for motion planning involving complicated robot dynamics. The efficiency and effectiveness of the proposed approach is shown by numerical results. Experimental results are used to show the feasibility of the presented planning algorithm.

Development of Application for Forest Insect Classification using CNN

Abstract: Insect identification has the disadvantage that it is difficult for non-experts to carry out due to the specificity of insects. Therefore, it is necessary for the general user to use auxiliary tools such as books to identify insects for education such as ecological learning. In recent years, researches using Deep Learning in fields such as object detection, behavior recognition, voice recognition as well as cancer detection in medical field have been actively conducted and show excellent results. In this paper, we developed a classification application that can be used in mobile phones with high automation and portability to solve the above insect classification problems. Experiments were conducted on 30 insect species selected for observable insects irrespective of environmental factors such as habitat and season, and the transform learning were applied to ResNet, which showed excellent performance in ILSVRC to classify forest insect. Our system achieved an average insect classification accuracy of 94%, an insect classification speed of 0.03 sec, and an insect photo transmission of 0.5 sec to output this information. This has proven that non-experts provide the appropriate performance to use.

Efficient Human-Robot Interaction using Deep Learning with Mask R-CNN: Detection, Recognition, Tracking and Segmentation

Abstract: We address social human-robot interaction problem by proposing an integration of deep neural network with mechanical robotic system to make it robust for human-robot interactive activities. Mask R-CNN, a neural network for object detection, can effectively help localize human faces which can be manipulated to instruct movement of the robot head. Our approach is not only suitable for detection and segmentation tasks but able to integrate as well with the mechanism of parallel mini-manipulator representing the 3D dimensions, in position and orientation of workspace. It can also solve the object segmentation problem which appears to be one of the most challenging issues in computer vision nowadays.

Interactive Analysis and Visualization of Digital Twins in High-Dimensional State Spaces

Abstract: Digital Twins (DTs), an emerging concept from Industry 4.0, are virtual representations of real technical assets. Multi-domain 3D simulation systems can bring DTs to life, even before their physical counterparts are finished. A DT's internal state can be fed from its real twin or generated by simulation. Access to this high-dimensional state of a DT is the key for various analysis and visualization methods presented in this paper. We introduce a generic formalism of state space for DTs and utilize it in an application scenario for automated driving. Throughout this example, methods for state logging and replays, data analysis, and visualization within 3D simulation frameworks are presented. Clear definitions for state variables, vectors, trajectories, and time series help slicing the DTs' state spaces of enormous dimensionality. The presented methodology does not only support the development of intelligent algorithms for autonomous driving, but is also the basis for further use cases of DTs involving optimization, mental models, and decision support systems.

An hybrid simulation tool for autonomous cars in very high traffic scenarios

Abstract: This article introduces an open source tool for simulating autonomous vehicles in complex, high traffic, scenarios The proposed approach consists on creating an hybrid simulation, which fully integrates and synchronizes two well known simulators: a microscopic, multi-modal traffic simulator and a complex 3D simulator The presented software tool allows to simulate an autonomous vehicle, including all its dynamics, sensors and control layers, in a scenario with a very high volume of traffic The hybrid simulation creates a bi-directional integration, meaning that, in the 3D simulator, the ego-vehicle sees and interacts with the rest of the vehicles, and at the same time, in the traffic simulator, all additional vehicles detect and react to the actions of the ego-vehicle Two interfaces, one for each simulator, where created to achieve the integration, they ensure the synchronization of the scenario, the state of all vehicles including the ego-vehicle, and the time The capabilities of the hybrid simulation was tested with different models for the ego-vehicle and almost 300 additional vehicles in a complex merge scenario

Vehicles Platooning in Urban Environment: Consensus-based Longitudinal Control with Limited Communications Capabilities

Abstract: In this research, a general control framework for platooning in urban environment is proposed A consensus-based control law is described taking into account the nature of traveling in urban environment, that is the human driven leader travels with variable velocity In addition, the proposed control law does not depend on the predecessor velocity, which in turn allows us to utilize a low cost limited bandwidth communication module by using a sensor-based link for predecessor distance and a communication-based link for leader's information A constant-spacing policy is used to get a high capacity flow of vehicles The control system is analyzed and conditions for both internal and string stability are set The efficiency of the proposed framework and control law is verified via numerical analysis

A Novel Warehouse Multi-Robot Automation System with Semi-Complete and Computationally Efficient Path Planning and Adaptive Genetic Task Allocation Algorithms

Abstract: We consider the problem of warehouse multi-robot automation system in discrete-time and discrete-space configuration with focus on the task allocation and conflict-free path planning We present a system design where a centralized server handles the task allocation and each robot performs local path planning distributively A genetic-based task allocation algorithm is firstly presented, with modification to enable heuristic learning A semi-complete potential field based local path planning algorithm is then proposed, named the recursive excitation/relaxation artificial potential field (RERAPF) A mathematical proof is also presented to show the semi-completeness of the RERAPF algorithm The main contribution of this paper is the modification of conventional artificial potential field (APF) to be semi-complete while computationally efficient, resolving the traditional issue of incompleteness Simulation results are also presented for performance evaluation of the proposed path planning algorithm and the overall system

Detection and Following of Moving Targets by an Indoor Mobile Robot using Microsoft Kinect and 2D Lidar Data

Abstract: The mobile robot following behavior is frequently considered as an important pre-requisite while developing autonomous service robots intended to co-exist with humans in a shared environment. It can also simplify the development of autonomous navigation and obstacle avoidance behavior as well as the robot ability to operate within multi-agent formations. In this investigation, an approach is proposed that allows a KUKA youBot omnidirectional mobile platform to detect and follow people carrying different objects, such as suitcases, in a shared indoor environment. Using a Kinect sensor, the mobile robot can recognize standing human with a suitcase and will begin to consider him as a potential dynamic target. The 2D lidar of the mobile platform can further detect when the above target starts to move and will begin to track it. Meanwhile, the robot will start following the human by controlling its own velocity while maintaining a pre-specified orientation and distance.

Continuous Gesture Recognition from sEMG Sensor Data with Recurrent Neural Networks and Adversarial Domain Adaptation

Abstract: Movement control of artificial limbs has made big advances in recent years. New sensor and control technology enhanced the functionality and usefulness of artificial limbs to the point that complex movements, such as grasping, can be performed to a limited extent. To date, the most successful results were achieved by applying recurrent neural networks (RNNs), However, in the domain of artificial hands, experiments so far were limited to non-mobile wrists, which significantly reduces the functionality of such prostheses. In this paper, for the first time, we present empirical results on gesture recognition with both mobile and non-mobile wrists. Furthermore, we demonstrate that recurrent neural networks with simple recurrent units (SRU) outperform regular RNNs in both cases in terms of gesture recognition accuracy, on data acquired by an arm band sensing electromagnetic signals from arm muscles (via surface electromyography or sEMG). Finally, we show that adding domain adaptation techniques to continuous gesture recognition with RNN improves the transfer ability between subjects, where a limb controller trained on data from one person is used for another person.

A Review on Short-Term Electricity Price Forecasting Techniques for Energy Markets

Abstract: The electricity price forecasting (EPF) is essential for decision-making mechanisms of market participants to survive in the deregulated and competing commercial environment. Due to special features of the electricity such as seasonality, the constant balance between production and consumption required by the system, and environmental dependencies, electricity prices generally shows extreme volatility and price spikes with the heteroscedasticity. This paper provides a survey of main EPF methodologies and the ultimate goal of this survey is to provide readers insights and guidelines for choosing different EPF techniques for day-ahead electricity markets. For each type of method, we briefly introduce its principle and then describe how it is applied in EPF. Many new EPF techniques developed recently are also discussed, especially the artificial intelligence forecasting methods. The pros and cons of each type of method are provided in a final table so that users can pay attention to when choosing them. In the final section, several promising methods and potential directions for further exploration are presented.

Post-Mission Autonomous Return and Precision Landing of UAV

Abstract: As recalling an Unmanned Aerial Vehicle (UAV) after completing a mission requires quite a lot of attention and skill from its operator, in this paper we propose a method to empower UAV with the capability to autonomously return to base and perform precision landing after completing a mission. The main challenge being tackled in this work is that while the vision-based landing technique is already mature, due to GPS error, UAV can only return to within several meters of home position after completing a mission and may fail to detect the visual marker. To resolve this problem, we employ Ultra-wideband (UWB) ranging measurements to localize and approach the home station. Once the UAV detects the visual marker, both UWB and visual tracking information are fused with onboard sensor to achieve even more accurate positioning. Real-life experiment is used to demonstrate the efficacy of the proposed scheme.

Building Occupancy Detection from Carbon-dioxide and Motion Sensors

Abstract: Occupant detection using carbon-dioxide sensors is prevalent but its accuracy is restricted by the inherent sensing delays. This paper proposes an indoor occupant detection method using real-time carbon-dioxide and Pyroelectric Infrared (PIR) sensor measurements overcoming the sensing delays. The occupancy detection problem is formulated as a classification problem wherein the classifier learns from offline carbon-dioxide data and the actual occupancy measurements of the room. While the classifier can provide realtime occupancy detection, the delays in carbon-dioxide sensors influence their accuracy. To overcome the delays, observations from PIR sensors are combined with the results of the single-layer feedforward neural network (SLFN) based classifier. The classifier works in four steps: (i) data-preprocessing, (ii) feature-selection, (iii) learning, and (iv) validation. The data is preprocessed by smoothing and several features are selected as input to the SLFN. Then, the classifier is validated with realtime experiments. Our results demonstrate that the proposed approach provides accuracy up to 99.79% and also overcomes the delays found in carbon-dioxide sensors.

An Autonomous Forklift with 3D Time-of-Flight Camera-Based Localization and Navigation

Abstract: In this paper we present an autonomous guided vehicle (AGV) based on a forklift. While centralized transport systems are widely used in the industry, these systems are expensive to set up and inflexible with regard to changes in the schedule. We in contrast investigate an approach that uses a decentralized autonomous forklift that uses a 3D Time-of-Flight (ToF) camera as its navigation sensor and is not dependent on artificial visual landmarks. The capability to process transport orders and maneuver in confined space with the required accuracy has been successfully tested in two warehouse environments.

Parameter Identification of Blimp Dynamics through Swinging Motion

Abstract: Indoor miniature autonomous blimp (MAB) is a small-sized aerial platform with outstanding safety and flight endurance. A detailed six-degree-of-freedom (6DOF) dynamics model is critical for controller design and motion simulation. This paper presents the identification of the rotation-related parameters of the blimp dynamics model through swing motion of the robot. A pendulum-like grey box model is constructed to identify the parameters from physical measurements and system identification experiments. The pendulum-like dynamics model with identified parameters is then linearized for future controller design and validated with experimental data.