Continuous Real-time Operation Research Articles

Machine learning based driver monitoring system: A case study for the Kayoola EVS

With the ever-growing traffic density, the number of road accidents has continued to increase. Finding solutions to reduce road accidents and improve traffic safety has become a top priority for Kiira Motors Corporation, a Ugandan state-owned automotive company. The company seeks to develop intelligent driver assistance systems for its market entry product, the Kayoola EVS bus. A machine learning-based driver monitoring system that would monitor driver drowsiness and send out an alarm in case drowsiness is detected has been developed in an attempt to reduce drowsiness-related accidents. The system consists of a camera positioned in such a way as to keep track of the driver's face. The camera is interfaced with a Raspberry Pi minicomputer which carries out the computations and analysis and when drowsiness is detected, an alarm is triggered. Dangerous driver behavior including distraction and fatigue has long been recognized as the main contributing factor in traffic accidents. This paper therefore presents the development of a driver monitoring system for the Kayoola Electric City Bus to address the increasing occurrences of road accidents. The machine learning-based driver monitoring system is designed to be non-intrusive with continuous real-time operation.

Dynamic average modelling of renewable generation sources for real time simulation

The increase in microgrids and smart grid research comprising renewable generation sources with advanced grid control functionalities is driving the need for sophisticated simulation and hardware in the loop (HIL) test facilities using digital real time simulators (DRTS). Unlike traditional off line simulation tools, DRTS provides fast, continuous real time operation as well as hardware interfaces to test the operation of physical control, protection and power devices before they are installed in the smart/micro grid application. The available computation resources of the DRTS to achieve fast, continuous real time operation impose a limitation on the size and level of detail of the power system model for the simulation and HIL application. Reduced dynamic models that represent the power system and control dynamics with sufficient accuracy provide an acceptable trade-off between the required computation resources and size of the power system model for large simulation applications. This paper discusses the dynamic average modelling of renewable energy generation sources for real time simulation applications.

Analysis of Machine Learning Methods in EtherCAT-Based Anomaly Detection

Today, the use of Ethernet-based protocols in industrial control systems (ICS) communications has led to the emergence of attacks based on information technology (IT) on supervisory control and data acquisition systems. In addition, the familiarity of Ethernet and TCP/IP protocols and the diversity and success of attacks on them raises security risks and cyber threats for ICS. This issue is compounded by the absence of encryption, authorization, and authentication mechanisms due to the development of industrial communications protocols only for performance purposes. Recent zero-day attacks, such as Triton, Stuxnet, Havex, Dragonfly, and Blackenergy, as well as the Ukraine cyber-attack, are possible because of the vulnerabilities of the systems; these attacksare carried by the protocols used in communication between PLC and I/O units or HMI and engineering stations. It is evident that there is a need for robust solutions that detect and prevent protocol-based cyber threats. In this paper, machine learning methods are evaluated for anomaly detection, particularly for EtherCAT-based ICS. To the best of the author's knowledge, there has been no research focusing on machine learning algorithms for anomaly detection of EtherCAT. Before testing anomaly detection, an EtherCAT-based water level control system testbed was developed. Then, a total of 16 events were generated in four categories and applied on the testbed. The dataset created was used for anomaly detection. The results showed that the k-nearest neighbors (k-NN) and support vector machine with genetic algorithm (SVM GA) models perform best among the 18 techniques applied. In addition to detecting anomalies, the methods are able to flag the attack types better than other techniques and are applicable in EtherCAT networks. Also, the dataset and events can be used for further studies since it is difficult to obtain data for ICS due to its critical infrastructure and continuous real-time operation.

Ultrasound flow mapping of complex liquid metal flows with spatial self-calibration

Investigating the complex interaction of electrically conductive fluids and magnetic fields is relevant for a variety of applications from basic research in magnetohydrodynamics (MHD) to modeling of industrial processes involving metal melts, such as steel casting and crystal growth. However, experimental studies in this field are often limited by the performance of flow instrumentation for opaque liquids. Commercially available measurement systems usually lack the ability to provide a time-resolved imaging of transient flow structures. We present an ultrasound array Doppler velocimeter (UADV) for flow mapping in opaque liquids near room temperature. It is modular and flexible regarding its measurement configuration, for instance it allows capturing two velocity components in two planes (2d–2c) of 67×67mm2 with a frame rate of 30Hz. It uses up to 9 linear arrays with 25 ultrasound transducers each driven in a parallelized time division multiplex (TDM) scheme. A FPGA-based signal processing allows a continuous realtime operation of the measurement system. Combining the single-component velocity data of each linear array to a 2d–2c flow field demands precise knowledge of the relative geometric position of the transducer arrays. We present a novel method that performs a spatial self-calibration by a mutual time of flight measurement, significantly reducing alignment errors. A measurement example of a magnetically stirred flow of GaInSn in a rectangular container is given. The UADV is applied to experiments in the context of manufacturing crystalline silicon ingots for photovoltaics.

Effect of pre-heating on microwave enhanced advanced oxidation process

Aerobic sewage sludge and dairy manure were pre-treated by conventional heating (CH) (40, 50 and 60°C) followed by microwave enhanced advanced oxidation process (MW/H2O2-AOP) at 100 and 120°C for solids disintegration and nutrient release. Increasing CH and MW temperatures resulted in an increase in soluble chemical oxygen demand concentration and volatile fatty acids in the treated solution. The maximum ratio of soluble to total chemical oxygen demand obtained for sewage sludge and dairy manure was 60% and 22%, respectively. The synergistic effect of microwave irradiation and hydrogen peroxide (H2O2) was more pronounced for sewage sludge at high CH and MW temperatures. Temperature and treatment time played a significant role in releasing orthophosphates. A maximum orthophosphate-to-total phosphate ratio of 55% and 85% was obtained for sewage sludge and dairy manure, respectively. A simple energy balance reveals that the CH step proves to be advantageous during continuous real-time operation of MW/H2O2-AOP technology.

Generating multimedia presentations that summarize the behavior of dynamic systems using a model-based approach

This article describes a knowledge-based method for generating multimedia descriptions that summarize the behavior of dynamic systems. We designed this method for users who monitor the behavior of a dynamic system with the help of sensor networks and make decisions according to prefixed management goals. Our method generates presentations using different modes such as text in natural language, 2D graphics and 3D animations. The method uses a qualitative representation of the dynamic system based on hierarchies of components and causal influences. The method includes an abstraction generator that uses the system representation to find and aggregate relevant data at an appropriate level of abstraction. In addition, the method includes a hierarchical planner to generate a presentation using a model with discourse patterns. Our method provides an efficient and flexible solution to generate concise and adapted multimedia presentations that summarize thousands of time series. It is general to be adapted to different dynamic systems with acceptable knowledge acquisition effort by reusing and adapting intuitive representations. We validated our method and evaluated its practical utility by developing several models for an application that worked in continuous real time operation for more than 1year, summarizing sensor data of a national hydrologic information system in Spain.

New Lightning Detection Networks in Poland – LINET and LLDN

Lightning detection in Poland is performed by means of a PERUN (Safir 3000) system operated by the Institute of Meteorology and Water Management. Poland is also partly covered by a VLF/LF lightning detection system (CLDN, Central Lightning Detection Network). Both sources of lightning data have their limitations resulting from detection technique, limited number of sensors and geographical configuration, with the consequence of shortcomings in the data quality. For this reason, a new network has been installed in Poland and started continuous real-time operation in May 2006. It is LINET that covers entire Poland and is complemented by numerous sensors positioned in surrounding countries. In 2007 additional LINET sensors have been installed in Poland in order to allow exploitation of reduced baselines for efficient achievement of total lightning. In the frame of the COST P18 Action “Physics of Lightning Flash and Its Effects” another new Polish project started in 2006 related to regional lightning location. At present, the Local Lightning Detection Network (LLDN) undergoes installation in the region of Warsaw. LLDN will consist of six individual stations equipped with E-field antennae and digital recorders synchronized with GPS time signals. The aim of LLDN installation is complement other networks covering region of Warsaw (PERUN, LINET) and to provide an additional source of lightning CG data with high sensitivity in a relatively small area. In the paper are described general characteristics of LINET in Poland, as well as basic characteristics and assumed performance of LLDN, which will start operation in 2008.

Security and Privacy in Distributed Smart Cameras

Distributed smart camera systems are becoming increasingly important in a wide range of applications. As they are often deployed in public space and/or our personal environment, they increasingly access and manipulate sensitive or private information. Their architectures need to address security and privacy issues appropriately, considering them from the inception of the overall system structure. In this paper, we present security and privacy issues of distributed smart camera systems. We describe security requirements, possible attacks, and common risks, analyzing issues at the node and at the network level and presenting available solutions. Although security issues of distributed smart cameras are analogous to networked embedded systems and sensor networks, emphasis is given to special requirements of smart camera networks, including privacy and continuous real-time operation.

Application of a flow type quartz crystal microbalance immunosensor for real time determination of cattle bovine ephemeral fever virus in liquid

Bovine ephemeral fever (BEF) is a viral disease of cattle. A flow type quartz crystal microbalance (QCM) immunosensor was developed for the real time determination BEF virus (BEFV) that is suitable for clinical point-case diagnosis. Self-assembled monolayer (SAM) of thiols and sulphides by the cystamine–glutaraldehyde method was used for the immobilization of BEFV monoclonal antibody on the gold surface of a quartz crystal microbalance (QCM). A positive correlation was found between the virus concentration and frequency changes ( R 2 = 0.9962) on this QCM system. The reproducible rates for the 50 and 10 μg/mL samples were 4 and 13.9%, respectively. There was no interference from non-specifically adsorbed phage. Using this flow type QCM immunosensor, BEFV could specifically be detected with sensitivity comparable to a conventional enzyme-linked immunosorbent assay (ELISA). The measurement could be obtained directly, within several minutes, rather than hours as required visualizing the results of ELISA. In addition, the observation of reproducible and constant changes after successive additions of BEFV suggests that a QCM immunosensor in a flow cell could be developed for automated or continuous real time operation.

Exploratory modelling for active vibration control

The design scheme called exploratory modelling for controller optimization (EMCO) is used to obtain simple robust controllers for complex plant dynamics in vibration control. The allowable gain error (AGE) function is used to quantify a frequency dependent bound of model gain errors which allow for both robust vibration control and the required stability robustness margin to be maintained. AGE lends itself to iterative modelling and controller redesign and the resulting scheme is outlined. For continuous realtime operation the internal model control (IMC) scheme provides self-excitation for re-identification to improve stability robustness.

Characterization of seismic phases-an automatic analyser for seismograms

SUMMARY An automatic analyser which characterizes, as well as detects, seismic phases can be a significant help in the interpretation of seismograms. Such an analyser has been constructed using concepts from work in artificial intelligence. A seismogram is recognized as a hierarchical structure of subunits and the features characterizing a phase are extracted via a structural analysis. The automatic analyser detects the onsets of phases, separates the relevant waveform segments, fits phase structures to standard wavelets, and generates a set of five parameters (‘a quintuple’) to characterize a detected phase. The whole analysis system requires just a few input parameters and can be accomplished swiftly enough to allow continuous real-time operation. The analysis procedure is adaptive, including updating of the estimate of the local frequency so that it can be applied to broad-band records with a wide range of frequency content. By coupling the analysis procedure with real-time filtering, very good results for phase detection and characterization can be achieved for weak distant events. The characterization of phase segments also provides useful information for the comparison of seismograms.

FFT computation with systolic arrays, a new architecture

The use of the Cooley-Tukey algorithm for computing the 1-d FFT lends itself to a particular matrix factorization which suggests direct implementation by linearly-connected systolic arrays. Here we present a new systolic architecture that embodies this algorithm. This implementation requires a smaller number of processors and a smaller number of memory cells than other recent implementations, as well as having all the advantages of systolic arrays. For the implementation of the decimation-in-frequency case, word-serial data input allows continuous real-time operation without the need of a serial-to-parallel conversion device. No control or data stream switching is necessary. Computer simulation of this architecture was done in the context of a 1024 point DFT with a fixed point processor, and CMOS processor implementation has started. >

An Iterative Cell Switch Design for Hybrid Redundancy

A marriage of N-modular redundancy (NMR) and standby sparing has resulted in a promising redundancy technique for protecting those portions of a fault-tolerant system whose continuous real-time operation is essential. This technique, known as hybrid redundancy, consists of N identical modules connected to a majority voter to form an NMR core. Disagreement detectors instruct a switch to replace any of the N modules that disagree with the majority consensus by a standby spare. The switch is essential to the operation of the hybrid redundancy scheme. The system reliability is a product of the reliabilities of the switch and the hybrid redundancy scheme assuming a perfect switch. To realize the demonstrated potential of the latter a highly reliable, thus simple, switch is required. An iterative cell switch is proposed and demonstrated to save at least 25 percent, and more than 80 percent in some instances, of the complexity of a switch design presented elsewhere in the literature. The use of threshold, rather than majority, voters is considered and shown to yield a simpler design in some cases. Three techniques for decreasing the propagation delay through the iterative cell switch are presented as well as a scheme to implement retry of failed modules. Finally, five different switch designs are compared on a cost and complexity basis.

Switch Complexity in Systems with Hybrid Redundancy

The combination of N-modular redundancy (NMR) and standby sparing has resulted in a promising redundancy technique for protecting those portions of a fault-tolerant system whose continuous real-time operation is essential. This technique, known as hybrid redundancy, uses N + Sp identical modules. N of these are connected to a majority voter to form an NMR core. The remaining Sp are used as standby spares. Disagreement detectors instruct a switch to replace with a standby spare any of the N modules that disagrees with the majority consensus. Since the switch and disagreement detector, as well as the modules, must function properly for the system to perform its designed task, the overall system reliability depends on the disagreement detector, switch, and hybrid reliabilities. Hence the overall reliability is a function of switch reliability. A highly reliable, thus simple, switch is desirable. First, strategies where every spare can be switched into every voter position (totally assigned) are considered. An optimal strategy is developed where the number of states in the switch is the criterion used for optimality. Formulas for the switch state count for the various strategies are derived and the strategies compared. Next, partially assigned switching strategies, strategies where every spare need not be capable of occupying every voter position, are examined. It is shown that designs where all spares are assigned to the same t + 2 of the 2t + 1 voter positions have as good reliability as the more complex totally assigned switching strategy.