The aptitude to identify the emotional states of others and response to exposed emotions is an important aspect of human social intelligence. Robots are expected to be prevalent in society to assist humans in various tasks. Human-robot interaction (HRI) is of critical importance in the assistive robotics sector. Smart digital assistants and assistive robots fail quite often when a request is not well defined verbally. When the assistant fails to provide services as desired, the person may exhibit an emotional response such as anger or frustration through expressions in their face and voice. It is critical that robots understand not only the language, but also human psychology. A novel affection-based perception architecture for cooperative HRIs is studied in this paper, where the agent is expected to recognize human emotional states, thus encourages a natural bonding between the human and the robotic artifact. We propose a method to close the loop using measured emotions to grade HRIs. This metric will be used as a reward mechanism to adjust the assistant's behavior adaptively. Emotion levels from users are detected through vision and speech inputs processed by deep neural networks (NNs). Negative emotions exhibit a change in performance until the user is satisfied.

World population growing in conjunction with the preference to live in the cities; make the city management a challenging issue. Traditional Cities with their common features will not be able to handle the human needs. As a result, smart city and its beneficial outcomes will attract a lot of attention recently. In fact, Smart city/community that is the field of collecting and processing data from so many different areas while making proper decisions and feeding to all parts of the system will be the future style of the cities. In other words, smart community is a complex big data problem which should combine different fields of research to make a unit environment. This brief survey will deal with all aspects of the smart communities. There will also be explained different categories of the smart communities in addition to their future challenges. In this work we try to introduce some aspects of the smart transportation in more detail.

The capability to perform identity reidentification in a crowd (e.g., from video feeds from a network of cameras, and social media platforms, such as Facebook and Instagram) efficiently and effectively is increasingly important, as evident in recent real-world events (e.g., terrorist attacks on places of mass gatherings in different countries). However, real-time reidentification in a network of cameras, such as those deployed in a smart city, and from other sources, such as social media platforms, remains a challenging task. In this paper, a new embedding algorithm pipeline is presented to extract and administrate the crowd-sourced facial image features (e.g., social media platforms and multicameras in a dense crowd, such as a stadium or airport). The proposed facial embedding is a privacy-aware parameterized function, which maps facial images to high-dimensional vectors in order to facilitate the identification and tracking of individuals. In other words, we are able to uniquely identify person(s) of interest, without the need to determine their true identity. To extract the facial embedding information in crowds, concurrent residual neural network (ResNet) embedding pipeline for each camera is proposed. Specifically, facial embedding feature vectors are generated in real-time by each camera using the proposed enhanced ResNet architecture, which is trained with vectorized-l2-loss function for face recognition. The multivariate kernel density estimation matching algorithm is then applied to facial embedding pipelines generated by cameras at the fog cloud for identity reidentification and security verification. This allows us to ensure the privacy of individuals captured by the camera without compromising on the capability for identity reidentification. Evaluations using mixed datasets in real-time demonstrate that our proposed approach achieves a 2.6% accuracy over other state-of-the-art approaches.

This paper discusses utilizing sparse autoencoders for building regression models in order to predict real-valued timeseries data. The focus of this research is on exploiting and learning from the determining features of continuous data via stacked autoencoders, thus increasing the prediction accuracy of regression method. Archi-tecture comprising different layers of sparse autoencod-ers, where each level of autoencoders are trained based on the standard (typical) method are implemented and analyzed. In order to enhance the accuracy of the typical model of training autoencoders, cascaded model that benefits from the fusion of low-and high-level features is proposed. The objective is to improve vehicular traffic flow forecasting, since this area is a research field in pro-gress that impacts daily life. The SGD algorithm at the top level of deep architectures serves as the regression method. Evaluations are based on the precision accuracy of the algorithms applied to forecasting the traffic flow of a location down a key highway using the historical traffic data of several locations ahead in the Twin Cities Metro area of Minneapolis.

Although there is no cure to date, Alzheimer's disease detection in early stages has a significant impact on the patient's life in terms of cost, the progress, and helping to plan in advance for an appropriate healthcare in the life ahead as well as providing clinical etiologies for further research. This paper discusses implementing a feature fusion method utilizing sparse and denoising autoencoders to reveal the stage of Alzheimer's disease. Four cohorts consisted of individuals with Alzheimer's disease, late mild cognitive impairment, early mild cognitive impairment, and normal control groups are classified using multinomial logistic regression fueled by the fusion of high-level and low-level features. The high-level features are extracted from the stacked autoencoders. The results show that feature fusion enhance the performance of typical autoencoders. However, the performance of feature fusion using denoising autoencoders is superior to that of the sparse training of autoencoders in terms of overall accuracy, precision, and recall.

Advanced traffic management systems (ATMS) are heavily depending on traffic flow or equivalent travel time estimation. The main goal of this paper is to accomplish two different algorithms to perform offline and online traffic flow forecasting. A multi-layer perceptron (MLP), which is trained on yearly data, is utilized for mid-term offline predictions. Principal components analysis (PCA) is employed to speed up the training process. This model also serves as a baseline. The stochastic gradient descent deploys online forecasting. Both algorithms predict the flow of a location down a Trunk highway (the target point) using the history of flow of several locations ahead of the target point in Twin Cities Metro area in Minneapolis.

The large-scale Systems of Systems (SoS) are there to improve the provision of societal services. These are driven by global drivers such as population growth, aging, food security, energy security and climate change. However, a considerable number of SoS projects resulted in failures, causing detrimental economic impact. This paper presents a set of preliminary results from an European Commission (EC) funded project called T-AREA-SoS (Trans-Atlantic Research and Education Agenda on Systems of Systems). The primary purpose of T-AREA-SoS is to provide input to EC's 80 billion Horizon 2020 research programme in formulating a Strategic Research Agenda for Systems of Systems Engineering (SoSE) that spans American and European activities. An initial set of prioritised SoSE research elements derived from an online survey are discussed from an economic and societal stance using examples from heterogeneous sectors such as transport, energy and manufacturing.

The System of Systems (SOS) technology is an advanced technology for Intelligent Systems that is developed with multiple intelligent systems. Recently, there has been a growing interest in a class of complex systems (robotic swarm as an example) whose constituents are themselves complex. Performance optimization, robustness and reliability among an emerging group of heterogeneous systems in order to realize a common goal have become the focus of various applications including military, security, aerospace, space, manufacturing, service industry, environmental systems, and disaster management, to name a few. In this paper, we discuss how Fuzzy Reinforcement Learning (FRL) can be used in SOS.

In this paper, a new extension of Bidimensional Empirical Mode Decomposition (BEMD) based on the median filtering is presented. The new scheme is compared with the traditional Empirical Mode Decomposition (EMD) and Fast and Adaptive Bidimensional EMD (FABEMD) techniques. The new scheme manifests to be superior to both of them in image enhancement and particularly in edge enhancement. The comparison is performed, objectively, using the measure of enhancement (EME) and also visually. Furthermore, we show that by applying different thresholds in stopping condition for each Intrinsic Mode Function (IMF), we can derive more accurate IMFs.

Motivated by example problems arising in energy, environment and service systems, the panelists will review those systems, man and cybernetics methods that are essential in dealing with these problems. The focus will be on new model-based and computational techniques, including approaches concerned with systems of systems, conflict resolution and adaptive real-time decision making. Although the organizers will initially serve as panel presenters, it is expected that they will also act as facilitators in an ensuing discussion among conference attendees who are cordially invited to participate in this specially organized panel.

Vanilla is the second most expensive spice worldwide. The high cost of vanilla has led to the problem of dangerous adulterated substitutes. Its high cost is attributed largely to the labor intensive hand pollination required where the melipona bee is not present. This article proposes a method of segmenting vanilla images intended for robotic control of a future automated pollination system. We present the specialization of a hypercomplex numbers based segmentation technique for vanilla flower recognition. The specialization overcomes much of the difficulty of differentiating green flowers from their similarly colored surroundings. Comparison is given to previous hypercomplex numbers based segmentation without the specialization.

In this paper, we have presented a GPS and sonar based area mapping and navigation scheme for a mobile robot. A mapping is achieved between the GPS space and the world coordinates of the mobile robot which enables us to generate direct motion commands for it. This mapping enables the robot to navigate among different GPS locations within the mapped area. The GPS data is extracted online to get the latitude and longitude information of a particular location. In the training phase, a 2-D axis transformation is used to relate local robot frame with the robot world coordinates and then the actual world coordinates are mapped from the GPS data using a RBFN (radial basis function network) based Neural Network. In the second phase, direct GPS data is used to get the mapping into the world coordinates of mobile robot using the trained network and the motion commands are generated accordingly. The physical placement of sonar devices, their ranging limits and beam opening angles are considered during navigation for possible collision detection and obstacle avoidance. This scheme is successfully implemented in real time with Pioneer mobile robot from ActivMedia Robotics and GPS receiver. The scheme is also tested in the simulation to justify its application in the real world.

The control of a swarm of underwater robots requires more than just a control algorithm, it requires a communications system. Underwater communications is difficult at the best of times and so large time delays and minimal information is a concern. The control system must be able to work on minimal and out of date information. The control system must also be able to control a large number of robots without a master control, a decentralized control approach. This paper describes one such control method.

We describe new methods that solve well-known problems associated with high throughput, robust communication in underwater environments. By robust, we imply simultaneous increases in enabled underwater communication data rates, range, and reliability. Our methods borrow concepts from advanced multi-carrier modulation using orthogonal frequency division multiplexing (OFDM), multiple input multiple output (MIMO) space time codes (e.g. Alamouti diversity), and zero carrier frequency ultra wideband (UWB) wireless. In addition, we discuss system optimization involving the elimination of the phase synthesizer and analog mixer through direct sampling of the RF antenna port.

In order to develop a robotic system of systems the robotic platforms must be designed and built. For this to happen, the type of application involved should be clear. Swarm robots need to be self contained and powered. They must also be self governing. Here the authors examine various applications and a prototype robot that may be useful in these scenarios.

Simulation provides a low cost method of initial testing of control for robotic swarms. The expansion of robotic swarms to heterogeneous environments drives the need to model cooperative operation in those environments. The Autonomous Control Engineering center at The University of Texas at San Antonio is investigating methods of simulation techniques and simulation environments. This paper presents results from adapting simulation tools for diverse environments.

This paper presents a system of systems approach to threat detection through integration of heterogeneous independently operable systems. The approach is presented on a realistic situation where a human-controlled base robot, swarm robot(s), and sensors work together to obtain a decision about a possible threat in the environment. The base robot is remotely operated by a human using a haptic control system. The swarm robot(s) are autonomous and can accept directives from the base robot. Finally, sensors directly communicate with (report to) the base robot. In this scenario, heterogeneous systems and human interact in a system of systems architecture. With the inclusion of human expert and sensor verification of swarm robots, the system can successfully perform the threat detection and reduce the false alarms. Finally, a system of systems simulation framework including a base robot, a swarm robot, and two sensors is presented in addition to an experimental evaluation of the proposed SoS architecture.

A modified probabilistic neural network (PNN) for brain tissue segmentation with magnetic resonance imaging (MRI) is proposed. In this approach, covariance matrices are used to replace the singular smoothing factor in the PNN's kernel function, and weighting factors are added in the pattern of summation layer. This weighted probabilistic neural network (WPNN) classifier can account for partial volume effects, which exist commonly in MRI, not only in the final result stage, but also in the modeling process. It adopts the self-organizing map (SOM) neural network to overly segment the input MR image, and yield reference vectors necessary for probabilistic density function (pdf) estimation. A supervised "soft" labeling mechanism based on Bayesian rule is developed, so that weighting factors can be generated along with corresponding SOM reference vectors. Tissue classification results from various algorithms are compared, and the effectiveness and robustness of the proposed approach are demonstrated.

Several important application areas that will dominate systems, man, and cybernetic (SMC) efforts for at least the next decade, together with the methods that will require further research and development in order to appropriately address these application domains, are considered herein. More specifically, four broad and pervasive system domains are examined: service systems, infrastructure and transportation systems, environmental and energy systems, and defense and space systems. Given the nature of these four application domains, a number of new systems [i.e., holistic-oriented, including system-of-systems (SoS)], man (i.e., decision-oriented, including decision informatics), and cybernetic (i.e., adaptive-oriented, including real-time control) methods are identified and their further development are discussed. Clearly, the IEEE Society on Systems, Man, and Cybernetics has a great future; its systems, man, and cybernetic methods are relevant for addressing challenging problems arising in system domains that are becoming dominant in this 21st century. However, the methods must be refined and expanded to meet the changing needs of the 21st century; from a system to a system-of-systems vision, from a disciplinary to a multidisciplinary outlook, from a mass production to a mass customization focus, from a steady state to a real-time perspective, and from an optimal to an adaptive approach.

The autonomous behavior of tiny, low powered programmable sensing devices often requires constant monitoring for accurate and reliable functioning and data delivery. The problem of monitoring these sensing devices is generally hard due to the dense deployment. It is common to divide such a large problem (or system) into sub-problems (or subsystems) and solve. In sensor networks, we divide the large deployment of sensors into clusters and monitor these groups of sensors or clusters. In this paper, we propose a novel methodology to analyze the critical parameters (bandwidth usage, network congestion, number of dead nodes, activity in the region, and overall energy (power) consumption) of a given network of sensing devices in order to increase the lifetime and reliability of the network.

In this paper, we propose a modified probabilistic neural network, and applied it to brain magnetic resonance (MR) image segmentation. Comparing to conventional probabilistic neural network (PNN), the so-called weighted probabilistic neural network (WPNN): 1) uses covariance matrices instead of singular smoothing factor in kernel Junction; 2) adds weighting factors in pattern to summation layer. The WPNN is able to take partial volume effect, which exists commonly in MR imaging, into account not only in the final resulting stage, but also in the modeling process. We adopt self-organizing map (SOM) neural network as the previous step of WPNN classifier, to quantize the input data. SOM yields reference vectors necessary for probabilistic density function estimation. A supervised "soft" labeling mechanism based on Bayesian algorithm is developed, such that weighting factors can be generated for each SOM reference vector. Tissue segmentation results from various algorithms are compared, and the effectiveness and robustness of the proposed approach are demonstrated.

Three individual experts have been developed based on extended auto associative neural networks (E-AANN), Kohonen self organizing maps (KSOM), and the radial basis function based clustering (RBFC) algorithms. An integrated method is proposed later to combine the set of individual experts managed by a gated experts algorithm, which assigns the experts based on their best performance regions. We have used a Matlab Simulink model of a chiller system and applied the individual experts and the integrated method to detect and recover sensor errors. It has been shown that the integrated method gets better performance in diagnostics and prognostics compared with each individual expert.

We present the modeling and control system's design of a laser pointing system. The aim of the paper is to provide means of detecting the relative errors of the system to correct them using actuating signals both directly and indirectly. The proposed control system will be used as part of a diagnostic and prognostic analysis for both systems. The paper includes the mathematical model of the two systems and Matlab simulations of the designed controller as well as its response to different stimulus. Diagnostic and prognostic approaches for actuator faults is presented.

This paper described the changes in the basic design of our mobile robot, ACE Rover I, but before that the original design of the ACE Rover I is discussed in brief. The proposed rover called ACE Rover I is a four-wheel drive, four-wheel steering, all-terrain autonomous rover. The first design had a lower level, which contained the wheel motors with encoders, a steering motor, tilt switch to determine surface inclination, bumpers to sense if the robot has bumped into an obstacle, an array of batteries, and four IR sensors to avoid abrupt surface changes such as drop-offs. The upper level has eight IR sensors to avoid large obstacles, an analog compass to determine direction, and the controller circuit board. The modified hardware design now has a sliding suspension system attached to the wheel unit, so that the robot can prevent mechanical failures due to shocks and vibrations in rugged terrain. With the introduction of the suspension system, we had to rethink the method of transmitting torque to each wheel. The prospect of adding a sonar ring or a vision system is also discussed in the paper. Towards the end of the paper, a brief description of previous approaches to localization and path planning with the available sensors on the ACE Rover I is given.

With the emerging applications of multi-agent systems, there is always a need for simulation to verify the results before actual implementation. Multi-agent simulation provides a test bed for several soft-computing algorithms like fuzzy logic, learning automata, evolutionary algorithms, etc. In this paper we discuss the fusion of these soft-computing methodologies and existing tools for discrete event simulation (DEVS) for multi-agent simulation. We propose a methodology for combining the agent-based architecture, discrete event system and soft-computing methods in the simulation of multi-agent robotics and network security system. We also define a framework called Virtual Laboratory (V-Lab/spl reg/) for multi-agent simulation using intelligent tools.

This paper presents a navigation methodology for decentralized autonomous automated guided vehicles used for material handling. The navigation methodology is based on behavior-based control augmented with multirobot coordination behaviors and a priori waypoint determination. Results indicate that the developed methodology fuses well between the desires for optimal vehicle routes on the one hand and decentralized reactive operation on the other.

In this paper, we present the use of stochastic learning automata (SLA) in multiagent robotics. In order to fully utilize and implement learning control algorithms in the control of multiagent robotics, an environment for simulation has to be first created. A virtual laboratory for simulation of autonomous agents, called V-Lab is described. The V-Lab architecture can incorporate various models of the environment as well as the agent being trained. A case study to demonstrate the use of SLA is presented.

In recent years, evolution-based knowledge optimization has gained a great deal of popularity due to its inherent ability in efficient and parallel search of complex and multimodal landscapes. Application of genetic algorithms (GA) to knowledge enhancement involves several aspects. First is how to code a string to represent all the necessary degrees of freedom for search in the fuzzy knowledge domain. The second aspect is how to incorporate existing expert knowledge into the GA-optimising algorithm. And in general, how to take advantage of several experts' opinions in creation of an initial population. Conventional applications of GA-fuzzy suggest using a random initial population. However, it is intuitively clear that any search routine could converge faster if starting points are good solutions. In this paper, a methodology is illustrated which incorporates expert knowledge in creating an initial population while allowing for randomness among members of the population for diversity. Furthermore, the methodology is applied to step response optimization of a flexible-link feedback control system.

A technique of identifying the dynamics of a robotics system using neural network is presented. The identified model is used by a fuzzy controller to evaluate the range of the control variables and also the performance of the adaptive control laws on the identified model. An overview of the neuro-fuzzy control architecture is also discussed. This architecture uses two neural networks, one which identifies the system dynamics and another classifies the temporal response of the robotic system. The information from the neural networks is used to make suitable adjustments in the parameter of the fuzzy controller. This paper however concentrates on the theory and operation of identifying the dynamics of a Adept-Two industrial robot. Simulation results are presented.

Adaptive behavioral capabilities are necessary for robust navigation in non-engineered environments. A control approach to adaptive behavior is described which exploits the approximate reasoning facility of fuzzy logic. In particular, a behavior-based architecture for hierarchical fuzzy control of mobile robots is presented. Its structure is described as well as mechanisms of control decision-making which give rise to adaptive behavior. Control decisions result from a consensus of recommendations offered only by behaviors that are applicable to current situations. Indoor navigation examples demonstrate the practicality of the approach and reveals characteristics of multiple behavior interaction.

A decentralized adaptive control is proposed to stabilize and track the nonlinear interconnected subsystems with unknown parameters. The adaptation of the controller gain is derived by using model reference adaptive control theory based on Lyapunov's direct method. The adaptive gains consist of sigma , proportional, and integral combination of the measured and reference values of the corresponding subsystem. The proposed control is applied to the joint control of a two-link robot manipulator, and the performance in computer simulation corresponds with what is expected in theoretical development.>

A hierarchical scheme for the optimal control of large-scale nonlinear time-delay systems is proposed. A linear approximation of the original system is defined about a nominal trajectory. The system is decomposed intoNsubsystems in such a way that the delay terms as well as interaction terms can be treated as extra inputs to linear nondelay subsystems. A numerical example illustrates this approach.