Search

This paper presents a piezoresistive pressure microsensor composed of a Silicon on Insulator (SOI) layer with sensing elements (a pressure-sensitive diaphragm and four piezoresistors) and a glass cap for hermetic package. The pressure under measurement bends the pressure-sensitive diaphragm, producing resistance changes of underlining piezoresistors. Numerical simulations were conducted, where key structure parameters of piezoresistors were optimized, producing a sensitivity of 18.61 mV/(V·MPa) and a linearity of 0.74%FS. The proposed microsensor was fabricated based on simplified fabrications, which included only two etching processes and one anodic bonding. Compared to other microfabrications, the fabrication of the developed microsensor does not need ion implantation and thin-film deposition, leading to high uniformity with low residual stress. Fabricated microsensors were characterized, obtaining the sensitivity of 17.278 mV/(V·MPa) and the linearity of 0.613 %FS (0-2.5 MPa).

We propose an efficient and robust segmentation network for automated seagrass region detection. The proposed network has a simple architecture to save computational demands as well as inference energy cost. More importantly, the scale of network can be feasibly adjusted, to balance the network computational demands and segmentation accuracy. Experimental results show that our proposed network is robust to segment the various seagrass patterns with 90.66% mIoU (mean Intersection over Union) accuracy. It had achieved 200 frames per second (FPS, 1.42 times faster than the second-best network GCN) on desktop GPU, and 18 FPS on NVIDIA Jetson TX2. It also has 3.45M parameters and 0.587 GMACs FLOPs (FLoating Point OPerations), only 14.6% and 10.8% of those in GCN respectively. To segment a single image on the Jetson TX2, our architecture requires an average energy of 0.26 Joule. This energy cost is only 46% of DeepLab, which shows the proposed network to be an energy efficient one. The proposed network demonstrates accurate and real-time segmentation capability, and it can be deployed to low-energy embedded AUVs for sea habitat protection.

Trust is essential in designing autonomous and semiautonomous Robots and Autonomous Systems (RAS), because of the “No trust, no use” concept. RAS should provide high quality services, with four key properties that make them trustworthy: they must be (i) robust with regards to any system health related issues, (ii) safe for any matters in their surrounding environments, (iii) secure against any threats from cyber spaces, and (iv) trusted for human-machine interaction. This article thoroughly analyses the challenges in implementing the trustworthy RAS in respects of the four properties, and addresses the power of AI in improving the trustworthiness of RAS. While we focus on the benefits that AI brings to human, we should realize the potential risks that could be caused by AI. This article introduces for the first time the set of key aspects of human-centered AI for RAS, which can serve as a cornerstone for implementing trustworthy RAS by design in the future.

Verification and validation of deep learning algorithms is an important and challenging topic of artificial intelligence. Without approving by reliable and rigorous verification methods, deep learning algorithms, for instance, the convolutional neural networks, are not qualified to be used in real-world applications, especially in safety-critical areas. The gap between deep learning systems and the requirements in safety-critical application areas, such as autonomous robotics and self-driving vehicles, is the lack of Black-box V&V techniques that can test and evaluate the performance and the robustness of deep learning systems. To bridge this gap, we proposed a GAN based Black-box verification framework called AdversarialStyle for generating and searching adversarial examples in both targeted and non-targeted way from different styles or domains of interest. The AdversarialStyle can not only evaluate deep learning models but also can discover the robustness level of every instance in the test set. Therefore, this framework can support deep learning model designers to understand and to explore their algorithms and improve the trustworthiness of AI techniques.

Herein, a combination of Local Pattern Generators (LPG) with reinforcement learning is proposed to balance a bipedal robot using minimal power consumption. This work presents the extension of Normalised Advantage Function (eNAF) algorithm to work with recurrent neural networks without sacrificing time-dependency between data in the same episode. Additionally, a hybrid controller is introduced by combining eNAF algorithm hierarchically with LPGs to provide more robustness with less computational power requirements. The system was asynchronous, as pattern generator had an activation frequency of 100Hz, while eNAF algorithm had only 1Hz and were not synchronised between them. Robot autonomy time was increased through reduction of computational load by introducing variable-ratio activation frequency between the LPGs and the eNAF algorithm. Finally, a new and novel bipedal robot design with non-conventional linear actuators was used as the basis of the simulator model. These experiments were implemented using V-Rep Edu simulator with the industrial Vortex Studio dynamic engine. The results demonstrate a fast and agile recovery by the trained robot after a push in transverse plane.

Maintaining safety in roundabouts is crucial in autonomous vehicles (AV) controlling and path planning. The number of vehicles in a roundabout at a time and the rules they must obey can make it a very complex traffic environment. Before an AV starts entering a roundabout, oncoming vehicles must be identified, and it could be done by determining their position, speed and direction. To address that, this paper extends some of our previous works in AV decision making in roundabouts and proposes a multi-grid-based image processing approach using multiple cameras (MGC). Particularly, it utilises a fine grid to determine speed and direction of approaching vehicles, whilst the position is evaluated using a larger grid. Besides, using multiple cameras allows the system to mimic the real drivers’ view and perception in approaching the real roundabouts, hence a human- like decision can be made. Three different classifiers including SVM, ANN and kNN were examined using 460 video clips of real roundabout-drive circumstances. The highest score was obtained by SVM at nearly 97% accuracy rate, with the making decision time is only around one second. That promising result indicates the applicability of the MGC system in real traffic situations.

Being able to join roundabouts correctly is crucial for an autonomous vehicle to maintain not only its own safety but also a normal traffic order for others. In order to know the right time and speed for entering roundabouts, the location, speed and direction of the approaching vehicles need to be taken into consideration. This study investigated the feasibility of leveraging computer vision and machine learning to help autonomous vehicles decide to wait or to enter when reaching roundabouts. A grid-based image processing approach with a single camera at normal roundabouts (GBIPA-SC-NR) is proposed in this paper to characterize traffic situations that can be used for machine learning algorithms to learn the roundabout joining criteria. Video road clips recorded when human drivers reach and then join various roundabouts at different locations were utilised for this learning process, with a selection of four supervised classification algorithms (i.e. the Support Vector Machines, Random Forests, K-Nearest Neighbours, and Decision Tree). The trained classifiers using the proposed approach were evaluated on 507 test videos captured at roundabouts, where the SVM showed the best performance with a 90.28% classification accuracy. This result suggests that the proposed grid-based image processing method can be applied to effectively help autonomous vehicles made the right decision when reaching a roundabout.

This paper1 presents the results of several machine learning techniques for step decision in a bipedal robot. The custom developed bipedal robot does not utilize electric motors as actuators and as a result has the disadvantage of imprecise movements. The robot is inherently unstable and maintain its stability by making steps. The classifiers had to learn when and which leg must be moved in order to maintain stability and locomotion. Methods like: Decision tree, Linear/Quadratic Discriminant, SVM, KNN and Neural Networks were trained. The results of their performance/accuracy are noted.

With the construction of global energy interconnection, plenty of fluctuating renewable energy access the distribution network, power system dynamic behavior is more complex. The wide-area measurement system light (WAMS Light) records the actual variation of the grid in the distribution network with high accuracy. Exploring the value of high-density, large-scale synchronous data and comprehensive visualization can efficiently reflect the behavior of the system. An open and interactive WAMS Light data visualization platform (WAMSLVP) is built with three-tier structure. WAMSLVP is designed by analysing the function requirements in multi-information fusion, multi-angle analysis, multi-layer integration, program flexible access based on the characteristics of distribution network. And WAMSLVP integrates the frequency, voltage magnitude, angle phase and other power information with the non-power information as multiple layers to compositive display based on geographic information system (GIS), facilitating the dispatchers simultaneously capture multiple information from different analysis just by a glance.

Robust real-time obstacle detection/avoidance is a challenging problem especially for micro and small aerial vehicles due to the limited number of the on-board sensors due to the battery constraint and low payload. Usually lightweight sensors such as CMOS camera are the best choice comparing with laser or radar sensors. For real-time applications, most studies focus on using stereo cameras to reconstruct a 3D model of the obstacles or to estimate their depth. Instead, in this paper, a method that mimics the human behavior of detecting the state of the approaching obstacles using single camera is proposed. During the flight, this method is able to detect the changes of the size area of the obstacles. First, the method detects the feature points of the obstacles, and then extracts the obstacles that has probability of getting close. In addition, by comparing the changes in the area ratios of the obstacle in the image sequence, the method can decide if it is obstacle or not. Finally, by estimating the obstacle 2D position in the image and combining with the tracked waypoints, the UAV can take the action of avoidance.

Unmanned Aerial Vehicles have been used widely in the commercial and surveillance use in the recent year. Vehicle tracking from aerial video is one of commonly used application. In this paper, a self-learning mechanism has been proposed for the vehicle tracking in real time. The main contribution of this paper is that the proposed system can automatic detect and track multiple vehicles with a self-learning process leading to enhance the tracking and detection accuracy. Two detection methods have been used for the detection. The Features from Accelerated Segment Test (FAST) with Histograms of Oriented Gradient (HoG) method and the HSV colour feature with Grey Level Cooccurrence Matrix (GLCM) method have been proposed for the vehicle detection. A Forward and Backward Tracking (FBT) mechanism has been employed for the vehicle tracking. The main purpose of this research is to increase the vehicle detection accuracy by using the tracking results and the learning process, which can monitor the detection and tracking performance by using their outputs. Videos captured from UAVs have been used to evaluate the performance of the proposed method. According to the results, the proposed learning system can increase the detection performance.

This paper describes enhancements made to the distributed performance impact (PI) algorithm and presents the results of trials that show how the work advances the state-of-the-art in single-task, single-robot, time-extended, multiagent task assignment for time-critical missions. The improvement boosts performance by integrating the architecture with additional action selection methods that increase the exploratory properties of the algorithm (either soft max or e-greedy task selection). It is demonstrated empirically that the average time taken to perform rescue tasks can reduce by up to 8% and solution of some problems that baseline PI cannot handle is enabled. Comparison with the consensus-based bundle algorithm (CBBA) also shows that both the baseline PI algorithm and the enhanced versions are superior. All test problems center around a team of heterogeneous, autonomous vehicles conducting rescue missions in a 3-dimensional environment, where a number of different tasks must be carried out in order to rescue a known number of victims that is always more than the number of available vehicles.

Rescue missions require both speed to meet strict time constraints and maximum use of resources. This study presents a Task Swap Allocation (TSA) algorithm that increases vehicle allocation with respect to the state-of-the-art consensus-based bundle algorithm and one of its extensions, while meeting time constraints. The novel idea is to enable an online reconfiguration of task allocation among distributed and networked vehicles. The proposed strategy reallocates tasks among vehicles to create feasible spaces for unallocated tasks, thereby optimizing the total number of allocated tasks. The algorithm is shown to be efficient with respect to previous methods because changes are made to a task list only once a suitable space in a schedule has been identified. Furthermore, the proposed TSA can be employed as an extension for other distributed task allocation algorithms with similar constraints to improve performance by escaping local optima and by reacting to dynamic environments.

Conference Name:2014 IEEE International Conference on Fuzzy Systems, FUZZ-IEEE 2014. Conference Address: Beijing, China. Time:July 6, 2014 - July 11, 2014.

Shepherding is often used in robotics and applied to various domains such as military in Unmanned Aerial Vehicle (UAV) or Unmanned Ground Vehicle (UGV) combat scenarios, disaster rescue and even in manufacturing. Generally, robot shepherding refers to a task of a robot known as shepherd or sheep herder, who guards and takes care of flocks of sheep, to make sure that the flock is intact and protect them from predators. In order to make an accurate decision, the shepherd needs to identify the flock that needs to be managed. How does the shepherd can precisely identify a group of animals as a flock? How can one actually judge a flock of sheep, is a flock? How does the shepherd decide how to approach or to steer the flock? These are the questions that relates to flock identification. In this paper, a new method using connected components labeling is proposed to cater the problem of flock identification in multi-robot shepherding scenarios. The results shows that it is a feasible approach, and can be used when integrated with the Player/Stage robotics simulation platform.

In recent years, unmanned aerial vehicles (UAVs) have gained a great importance in both military and civilian applications. In this paper, we proposed a vehicle detection method from UAVs which integrated of Scalar Invariant Feature Transform (SIFT) and Implicit Shape Model (ISM). Firstly, a set of key points was detected in the testing image by using SIFT. Secondly, feature descriptors around the key points were generated by using the ISM. Support Vector Machines (SVMs) were applied during the key points selection. The experiment used a video shoot by a UAV in a highway and the results showed the performance and the effectiveness of the method.

Motion segmentation plays an important role in human motion analysis. Understanding the intrinsic features of human activities represents a challenge for modern science. Current solutions usually involve computationally demanding processing and achieve the best results using expensive, intrusive motion capture devices. In this paper, a simple, affordable and effective method for human motion segmentation and alignment is presented. The approach follows a two-step process: first, the most salient principal components are calculated in order to reduce the dimension of the input motion data. Then, candidates of key body poses (landmarks) are inferred using multi-class, supervised machine learning techniques from a set of training samples. Finally, cluster analysis is used to refine the result. Predictions are guaranteed to be invariant to the repetitiveness and symmetry of the performance. Results show the effectiveness of the proposed approach by comparing it against the Dynamic Time Warping algorithm and Hierarchical Aligned Cluster Analysis.

This paper addresses the problem of multi-agent, multi-task assignment with multiple agent requirements on tasks for unmanned aerial vehicles by presenting the Consensus Based Grouping Algorithm. The algorithm is an extension of the Consensus Based Bundle Algorithm that converges to a conflict free, feasible solution of which previous algorithms are unable to account for. Furthermore the algorithm creates a framework to take into account task based requirements, deadlocking and a method to store assignments for a dynamical environment.

Cognitive distraction is happened when a driver's mind is off the road. It happened when a driver is looking on the road but his mind is doing a thinking process. It has been found that, cognitive distraction is the most dangerous type of driver distractions. This has been presented in the comparison table and stem plot between Control Experiment result and Task Experiment result. Information from eye movement and mouth movement are obtained using the faceLab cameras and their correlation is discussed here. Two sets of experiment (Control and Task) with 6 participants were completed for this paper. Results were presented in scatter diagram to show the correlation between eye and mouth movements. Stem plot is to show the different result obtained between control and task experiment.

Citrus canker is a quarantine disease which may cause huge damage to citrus production. Effective and fast disease detection methods must be undertaken to minimize the losses of citrus canker infection. In this paper, a new approach is presented to detect citrus canker from leaf images collected in field. Firstly, a global canker lesion descriptor is used to detect citrus diseased-lesion from leaf-background. Then a zone-based combined local descriptor is proposed to identify citrus canker disease from other similar diseased-lesions. Thirdly, a two-level hierarchical detection structure is developed to identify the canker lesion and AdaBoost is adopted in feature selection and classifier learning. Finally, evaluation of the proposed method and its comparison with other approaches are discussed, and the experimental results shows that the proposed approach achieves similar classification accuracy of human experts.

In this paper, basic biological immune systems and their responses to external elements to maintain an organism's health state are described. The relationship between immune systems and multi-robot systems are also discussed. The proposed algorithm is based on immune network theories that have many similarities with the multi-robot systems domain. The paper describes a refinement of the memory-based immune network that enhances a robot's action-selection process. The refined model; which is based on the Immune Network T-cell-regulated—with Memory (INT-M) model; is applied onto the dog and sheep scenario. The refinements involves the low-level behaviors of the robot dogs, namely Shepherds' Formation and Shepherds' Approach. The shepherds would form a line behind the group of sheep and also obey a safe zone of each sheep, thus achieving better control of the flock. Simulation experiments are conducted on the Player/Stage platform.

Driver distractions can be categorized into 3 major parts:-visual, cognitive and manual. Visual and manual distraction on a driver can be physically detected. However, assessing cognitive distraction is difficult since it is more of an “internal” distraction rather than any easily measured “external” distraction. There are several methods available that can be used to detect cognitive driver distraction. Physiological measurements, performance measures (primary and secondary tasks) and rating scales are some of the well-known measures to detect cognitive distraction. This study focused on physiological measurements, specifically on a driver's eye and mouth movements. Six different participants were involved in our experiment. The duration of the experiment was 8 minutes and 49 seconds for each participant. Eye and mouth movements were obtained using the FaceLab Seeing Machine cameras and their magnitude of the r-values were found more than 60% thus proving that they are strongly correlated to each other.

There are basically 3 different types of driver distraction: manual, visual and cognitive. This paper focused on cognitive distraction on drivers. Cognitive distraction is occurred when a driver's mind is off from the road. Drivers are might probably see and realize objects and the environment on the road, and manually handle their vehicle safely, but their minds are thinking something that are not related to a driving safety issue. Cognitive distraction is quite difficult to be detected compare to manual and visual distraction, thus, it is the most dangerous type of distraction. This paper is to compare men and women distractions cognitively. A physiological measurement specifically on eyes has been used as the feature to detect driver's cognitive distraction. Information on blinking frequency, blinking duration, gaze rotation and pupil diameter has been captured using faceLab Seeing Machine cameras. Data were analyzed using independent sample t-test and means from each men and women have been presented in bar graphs.

In this paper, basic biological immune systems and their responses to external elements to maintain an organism's health state are described. The relationship between immune systems and multi-robot systems are also discussed. Our proposed algorithm is based on immune network theories that have many similarities with the multi-robot systems domain. The paper describes a memory-based immune network that enhance a robot's action-selection process and can obtain an overall a quick group response. The algorithm which is named as Immune Network T-cell-regulated — with Memory (INT-M) is applied to the dog and sheep scenario. Simulation experiments were conducted on the Player/Stage platform and experimental results are presented.

The rapid growth of elderly population has created a huge increase in demand for home health-care services. Home service robots are expected to play an important role to support elderly people at home to improve their life quality. Unlike industrial robots, home service robots face many challenges in terms of safety, autonomy, flexibility and usability, etc. How to efficiently realize human-robot interaction is one of the main areas to achieve home service tasks. Understanding the mechanisms of empathy in human beings and applying such mechanisms to home service robots is crucial to make human-robot interaction natural and friendly. Recent studies shows the links in the functional, and neural mechanisms between imitation and empathy in human beings. In this paper, algorithms are developed to allow robots to understand natural color names given by the user to help locate target objects, and showing by examples is applied to express human's intentions to robots in imitation. Case studies are conducted to demonstrate the proposed approaches.

Advanced autonomous artificial systems will need incremental learning and adaptive abilities similar to those seen in humans. Knowledge from biology, psychology and neuro-science is now inspiring new approaches for systems that have sensory-motor capabilities and operate in complex environments. Eye/hand coordination is an important cross-modal cognitive function, and is also typical of many of the other coordinations that must be involved in the control and operation of embodied intelligent systems. This paper examines a biologically inspired approach for incrementally constructing compact mapping networks for eye/hand coordination. We present a simplified node-decoupled extended kalman filter for radial basis function networks, and compare this with other learning algorithms. An experimental system consisting of a robot arm and a pan-and-tilt head with a color camera is used to produce results and test the algorithms in this paper. We also present three approaches for adapting to structural changes during eye/hand coordination tasks, and the robustness of the algorithms under noise are investigated. The learning and adaptation approaches in this paper have similarities with current ideas about neural growth in the brains of infants during early cognitive development.

Range images of complex geometry presented by large point data sets almost always yield surface reconstruction imperfections. We propose a novel compact and complete mesh representation for non-uniformly sampled noisy range image data using an adaptive radial basis function network. The network is established using a heuristic learning strategy. Neurons can be inserted, removed or updated iteratively, adapting to the complexity and distribution of the underlying data. This flexibility is particularly suited to highly variable spatial frequencies, and is conducive to data compression with network representations. Experiments confirm the performance advantages of the network when applied to 3D point-cloud surface reconstruction.

In order to control the continuous passive motion (CPM) machine for injured fingers, we develop an embedded control platform. We first bring forward the philosophy of function modularization design for control platforms. Then we actually begin to develop an embedded control platform of the CPM machine by using the method of function modularization. The core of the control platform consists of two main parts: the data acquisition function module and the motor control function module, both are based on the serial peripheral interface (SPI) network. The whole control platform is open-ended for new functions and applications. It can be easily expanded if we add new modules to the SPI network. Primary experiments have proved that the control platform works well and the design method of function modularization provides a new method for the design of control platforms.

During early robot learning, several mappings need to be set up for sensorimotor coordinations and transformation of sensory information from one modality to another. Usually these mappings are nonlinear and traditional passive learning approaches can not deal with these problems well. In this paper, A hierarchical clustering technique is introduced to group large mapping error locations and these error clusters drive the system to actively explore details of these clusters. Higher level local growing radial basis function subnetworks are used to approximate the mapping residual errors from previous mapping levels. Plastic radial basis function networks construct the substrate of the learning system and a simplified node-decoupled extended kalman filter algorithm is presented to train these radial basis function networks. Experimental results are given to compare the performance between active learning and passive learning.

We describe an approach to sensory-motor learning and coordination that draws from psychology rather than neuroscience. The growth of the motor coordination is controlled through sequential lifting of constraints, which is inspired by Jean Piaget's developmental learning theory. Our objective is the implementation of a flexible learning framework for an embodied hand/eye system which exhibits a prolonged epigenetic developmental process. The results show how staged competence can be shaped by qualitative behavior changes produced by anatomical, computational and maturational constraints.

The worldwide population of elderly people is rapidly growing and is set to become a major problem in the coming decades. This phenomenon has the potential to create a huge market for domestic service robots that can assist with the care and support of the elderly. Robots that are able to help the user with specific physical tasks are likely to become very important in the future, but so far, unlike industrial robots, assistive robots are still under-developed and are not widely used. We analyse the nature of the requirements for assistive robotics for the elderly and argue that traditional "industrial" robot design and control approaches are inappropriate to tackle the key problem areas of safety, adaptivity, long-term autonomy of operation, user-friendliness and low costs. We present a novel approach to the control of autonomous assistive robots for the home, with emphasis on the special requirements for in situ learning, including software compensation for low precision hardware components. Our system consists of a modified behaviour-based architecture with integrated knowledge representation and planning abilities. Automatic error-recovery is implemented as an activation spreading mechanism and is distributed across the behaviour repertoire. Context-based experience is learned during both error recovery and normal action and assimilated into the behaviours. This allows reuse across different tasks, and facilitates gradual but life-long improvements in system performance. To evaluate our approach, an experimental laboratory testbed was constructed using low-cost, low-precision components. Our system was implemented in software and a series of experiments were performed in order to investigate a range of tasks. The tasks were selected to face some of the key issues identified and the results show the potential for software solutions to overcome the barriers to successful assistive robotics for the elderly. The methods, experiments and results are described in this paper.

This paper presents ongoing research on assistive service robotics for health-care and home applications. The whole control system is based on a behavior-based architecture that integrates reactive behavior into a task driven structure. Efficient and natural robot communications are promoted by employing the user's perceived object color names. This information is then integrated with simple object shape features for object location and recognition. An error recovery strategy is embedded in each behavior by incorporating reasoning and an error recovery template into the behavior structure. Error recovery experience is also stored in the template and used to adjust the error recovery action generation process to improve the system's performance during its lifetime.