In large-scale systems, the components are interconnected and so the variables are correlated, which constitutes information on system topology with causality. After a fault occurs, it not only shows up as local phenomenon but also propagates to some other components or variables. Hence we should consider the sensor location problem to find the root cause of the fault origin and type from the viewpoint of the whole system.In order to measure the fault detection quality related to sensor location, some criteria are defined in Kawabata et al.’s paper [1]. Firstly, all the faults should be detected when they occur. Secondly, different faults should be identified from each other so that one can differentiate them based on the sensor readings.

The criteria of detectability and identifiability are basic requirements for fault detection [2].

In this reference all the sensors are assumed to be effective, that is, they show exactly whether the process variables are normal or abnormal.In engineering practice, sensors may often be faulty, meaning that they may fail to give adequate readings. For example, the reading may remain unchanged when the true value should be a deviation, which is called a missed alarm; or the sensor may give an alarm for a normal operation state, known as a false alarm. We should therefore allow for some redundancy in sensors in case of failures. More commonly, the measurements may show Drug_discovery these two kinds of sensor faults because of the choice of the threshold.

Often due to noise there are no real Entinostat sensor faults but deviations due to measurement noise, which is inevitable.

If the threshold setting is strict in order to suppress the missed alarm probability, the reading will be sensitive to random noise and temporary deviations, resulting in a high probability of false alarm. If we relax the threshold and accept larger region to be considered as normal, then the number of false alarms will decrease with more missed alarms. Therefore, missed alarms and false alarms are two aspects of reliability and we have to make a trade-off between them. This can be clearly illustrated via a receiver operating characteristics (ROC) curve [3,4].

Sensors in this present paper also include soft sensors that measure some specific variables by soft sensing techniques [5].With increasing complexity in process industrial systems, traditional mathematical models are difficult to obtain. Hence, graph-based models are proposed in the modeling analysis. Based on the signed directed graph (SDG) model, Raghuraj, et al. [2] have discussed the problems of detectability and identifiability in sensor location and presented the corresponding algorithm for locating each sensor.

The SIFT detector has been successfully applied to various applications such as face recognition [33] and medical image registration [24�C27]. In image registration, the SIFT keypoints [24, 27] localized at specific anatomical structures could be automatically selected for the adaptive setup of the irregular control point grids for the local deformation of specific anatomical structures. Without tedious manual selection of control points, the adaptive setup of irregular control point grids could alleviate the computational cost and the registration inaccuracy that are related to the regular grids [34] of control points arranged for the local large deformation at the tumor resection region.

In general, it is often difficult to correctly match local keypoints [26, 28] by using only the similarity between SIFT descriptors when complex local deformation and outliers exist in brain images.

The deformation invariant local feature descriptor was presented in [29], however this topic is beyond the scope of this paper.Although the results of these above methods clearly demonstrate the power of local invariant feature-based nonrigid deformations, the desired landmark-based registration algorithm should establish robust control point correspondence to accurately model the complex local deformation around the tumor resection region. To find robust point correspondence, some approaches are proposed including soft correspondence detections [20, 35], joint clustering-matching strategy [36] and modeling point sets by kernel density function [37].

Compared with the classical template matching, the iterative closest point [38] and the correspondence by sensitivity to movement [39], the self-organizing map [40] algorithm was considered in [41] to be the most effective method in 2D feature point correspondence detection. However, these methods do not consider the complexity of correspondence detection in the context of local large structure distortion combined with the outliers. In this work, we first compute the global correspondence between the contiguous matching areas of normal tissues and the tumor regions in the two images by using MI-based rigid registration method.

The global correspondence is then used to introduce Dacomitinib the cluster correspondence between Cilengitide the paired pools of DoG keypoints [23] that are detected and clustered at those contiguous matching areas in the two images. An important contribution is that we have proposed the cluster-to-cluster correspondence can be introduced as a useful constraint for the local point correspondence detection within the paired pools of keypoints.

gh biosynthetic activity in grain for mation when the total dry matter starts to increase and endosperm starch begins to accumulate rapidly in the seed, whereas dur ing the latter phase the grain usually exhibits a slower increase in dry weight until maximum values are reached and grain weight becomes constant. Global gene expression profiling studies of mRNAs have shown that many genes in multiple pathways participate in grain filling processes, such as those involved in nutri ent synthesis, starch synthesis and transport. On the other hand, miRNAs were identified as prefer entially expressed in various rice organs, including leaf, root, panicle and stem, as well as in seedlings under various stress treatments. A number of studies were also carried out on small RNAs in the grains of ja ponica varieties.

Some miRNAs were preferen tially expressed in early developing rice grains, such as 1 10 DAF and 3 12 DAF, suggesting regulatory roles of miRNAs during grain development. These stud ies, mainly in subspecies of japonica, also identified sig nificant numbers of both conserved and non conserved miRNAs. We report here the generation and sequencing of a small RNA library from grain tissues sampled dur ing the entire grain filling stage of an indica cultivar. In addition to numerous conserved miRNAs, we identified 11 novel miRNAs. Subsequently, Dacomitinib a customized miRNA chip was generated and miRNA expression profiling was studied using RNA samples from grains of each of the three filling stages, viz. milk ripe, soft dough, and hard dough.

Our results showed that most of the widely conserved miRNAs were down regulated during grain develop ment whereas rice or grass specific miRNAs were up regulated. The targets of differentially expressed miRNAs appeared to be involved in multiple biological processes, such as carbohydrate metabolism, hormone signaling and pathways associated with seed maturity, suggesting that rice miRNAs may play important roles during grain development. Results Small RNA populations at the grain filling stage We measured the fresh and dry grain weights of rice cultivar, Baifeng B, an indica landrace, at several stages of grain filling. The fresh weights began to increase from 3 DAF, dry matter accu mulation became faster from 5 DAF and reached highest levels at about 25 DAF.

Morphological observations of developing rice seeds showed that the filling phase can be divided into three continuous filling stages. For Illumina sequencing, we isolated small RNAs from immature rice grains sampled at 5 DAF to 25 DAF. After removing low quality reads, a total of 1,832,288 clean reads were obtained with 974,934 unique sequences. About 637,362 distinct reads were aligned to the 9311 genome using short oligonucleotide alignment pro gram. Among them, 21 nt and 24 nt small RNAs form the two largest groups, accounting for 22. 3% and 50. 5% of raw reads, respectively. By comparison with miRNAs from miRBase v16. 0, 102 known miRNAs were found in our dataset

nvestigation. The endogenous e pression of podoplanin on 293T cells and the specific interaction of podoplanin with CLEC 2 raised the questions if podoplanin was incorpo rated into virions produced in 293T cells, and if incorpo ration of podoplanin was required for CLEC 2 binding of these virions. Western blot analysis and knock down of podoplanin e pression by shRNA provided affirmative answers to both questions Podoplanin depletion reduced CLEC 2, but not DC SIGN, dependent HIV 1 transmis sion by B THP cells, and diminished transmission by platelets by about 50%. The latter finding is in agreement with our previous observation that CLEC 2 specific antiserum reduced HIV 1 transmission by plate lets by about half.

Podoplanin therefore joins the list of host factors which can be incorporated into the HIV 1 envelope and impact HIV 1 infection by interacting with their cognate ligands. A prominent e ample for such a factor is ICAM 1 which was found to be incorpo rated into the viral membrane, and to facilitate HIV 1 infection by binding to its ligand LFA 1 on T cells. The potential relevance Anacetrapib of podoplanin incorporation for HIV spread in infected individuals is critically deter mined by the overlap of the podoplanin e pression pat tern with the cellular tropism of HIV. Analysis of T cell lines and PBMCs for podoplanin e pression yielded neg ative results, at least when viable cells were ana lyzed, indicating that HIV particles generated in patients might not harbour podoplanin. The e ception might be viruses released from kidney podocytes which have been documented to e press podoplanin and to be susceptible to HIV infection.

However, the biolog ical relevance of this process is questionable. In this con te t, it also needs to be noted that podoplanin e pression is up regulated in many tumours including Kaposi sar coma. Podoplanin CLEC 2 dependent platelet stimulation by tumour cells promotes hematogenous tumour metastasis, possibly by inducing growth factor secretion by platelets and by promoting formation of a platelet cap, which protects the tumour from mechanical forces. Thus, podoplanin might play a role in the development of the AIDS associated Kaposi sarcoma, but is unlikely to modulate HIV spread in patients.

Nev ertheless, HIV 1 produced in PBMCs was transmitted to target cells in a CLEC 2 dependent fashion, sug gesting that primary T cells might e press a so far unrec ognized CLEC 2 ligand, which is incorporated into the viral envelope and which facilitates HIV transmission by CLEC 2. Our ongoing studies are devoted to the identification of this factor. Podoplanin was not detected on viable CEM��174 cells and PBMCs, as determined by our gat ing strategy and by co staining with the apoptosis and necrosis markers anne in V and 7 AAD, respectively. In contrast, we observed efficient reactivity of two different podoplanin antibodies with non viable cells, raising the intriguing possibility that podoplanin might be e pressed at the cell surface i

Therefore, it is a promising research direction to combine the advantages of different sensing sources, because each modality has complementary benefits and drawbacks, as has been shown in other works Bellotto and Hu [5], St-Laurent et al. [6], M. Hofmann and Rigoll [7], Johnson and Bajcsy [8], Zin et al. [9].Additional requirements for our application arise from the fact that the low-cost thermal sensor provides a low resolution image and, therefore, does not allow us to build accurate models for detecting people. Moreover, in order to have a high reaction capability, we are looking for solutions that allow parallel processing of all the input data instead of sequentially.

Therefore, the chosen approach is:To combine machine learning paradigms with computer vision techniques in order to perform image classification: first, we apply transformations using computer vision techniques, and second, we perform classification using machine learning paradigms.To construct a hierarchical classifier combining the three sensor source data (images) to improve person detection accuracy.We have evaluated the system in two different real scenarios: a manufacturing shop floor, where machines and humans share the space while performing production activities, and a science museum with different elements exposed, people moving around and strong illumination changes, due to weather conditions. Experimental results seem promising considering that the percentage of wrong classifications using only Kinect-based detection algorithms is drastically reduced.

The rest of the paper is organized as follows: In Section II, related work in the area of human detection is presented. We concentrate mainly on work done using machine learning for people detection. Section III describes the proposed approach and Section IV, the experimental evaluation. Section V shows experimental results and Section VI, conclusions and future work.2.?Related WorkPeople detection and tracking systems have been studied extensively because of the increasing demand for advanced robots that must integrate natural human-robot interaction (HRI) capabilities to perform some specific tasks for the humans or in collaboration with them. A complete review on people detection is beyond the scope of this work; extensive work can be found in Schiele [10] and Cielniak Drug_discovery [11]. We focus on the recent related work.To our knowledge, two approaches are commonly used for detecting people on a mobile robot: (1) vision-based techniques; and (2) combining vision with other modalities, normally range sensors, such as laser scanners or sonars, like in Wilhelm et al. [12], Scheutz et al. [13], Martin et al. [14]. Martin et al.

As we will analyze, these simplifications still comply with the important parts of both standards to facilitate a real deployment. We also discuss a small testbed, which we have deployed to obtain processing times and message sizes that lead to important conclusions about the usage of PANA in networks of constrained devices. To extend the analysis, we have used Cooja [11] to run simulations with several nodes.The remainder of the article is organized as follows. Section 2 presents some related work, and Section 3 presents some important background for understanding our implementation and the corresponding results. In particular, PANA and EAP are described, as well as the most relevant aspects of the protocol, such as the associated architecture.

Section 4 identifies important design decisions, which we have taken to adapt PANA and EAP to constrained devices without greatly affecting the standards. Section 5 provides some results obtained from a testbed especially designed to evaluate our implementation. Finally, we provide some conclusions and future work guidelines in Section 6.2.?Related WorkThe network access control and bootstrapping procedures in constrained devices are important topics nowadays. The authors in [12] expose the main features of IP-based security protocols for bootstrapping. It is shown that, in general, security protocols used today on the Internet were initially designed for nodes with high computational capabilities and permanent power supply, a large amount of memory and network links with sufficient bandwidth. However, this is not the case in constrained devices.

GSK-3 The capabilities of these devices are much lower than the general purpose ones. Furthermore, the programming paradigm and mode of operation of this type of network change.In [13], the authors give a complete overview of the security bootstrapping solutions for constrained devices. Five areas of bootstrapping are defined: user interface, bootstrap profile, security method, bootstrap protocol and communication channel. The user interface provides the interaction between the user and the bootstrap protocol. The user interface will vary depending on the capabilities on the node. In most cases, the user interface does not exist, and all the parameters needed by the bootstrap protocol are configured statically. Those parameters are saved in the bootstrap profile, which defines what information should be exchanged during the bootstrapping process. Potentially, a single node may run the protocol multiple times with different profiles, although they should be previously defined. The security method defines supported mechanisms for bootstrapping.

05 ��m. The sensing length of the FBGs was about 10 mm. The reflectivity of the resulting FBG was about 99%, and the peak wavelengths were between 1,550 and 1,551 nm. The full width half maximum (FWHM) of the FBGs was about 0.175 nm. Impacts were made at either of the two locations designated A and B in Figure 1b, using a 260 g aluminum weight falling from a height of 140 cm with an apparatus that conforms to ASTM D5628. B was the position of the FBG and position A was 30 mm from B. The fiber on the side that faced the impact was designated L1 and the one on the back surface L4. After impact the coupons were subjected to cyclic fatigue loading from 0.5 to 5 kN at a frequency of 5 Hz using an MTS servo-hydraulic testing system 810 (MTS Systems Corporation, Eden Prairie, MN, USA) for 200,000 cycles.

The reflected spectra from the FBGs were interrogated periodically using an optical spectrum analyzer (Anritsu MS9710C OSA, Anritsu Company Ltd., Kanagawa, Japan) under the load-free condition. The above tests have also been repeated on specimens without undergoing any impact to serve as control. Ultrasonic C-scan w
Inductive position sensors are widely used in modern automotive and industrial applications [1�C3]. They have various benefits such as low cost, good insensitiveness against temperature, and no wear-out [4�C6]. Several types of position sensors based on the inductive principle differ in their nonlinearity errors [7]. The grating eddy current position sensor not only has the function of resisting liquids, but it also prevents ferromagnetic particles from affecting measurement results.

However, measurement blind areas are not completely eliminated, so the linearity of the sensor is not satisfactory [8,9]. Inductive angle sensors are not susceptible to background electromagnetic interference, and they produce much greater output signal levels compared to other choices. However, there are usually higher order harmonic signals which lead to a considerable amount of nonlinearity AV-951 errors [10]. Inductive angle sensors provide a compact structure and a high degree of insensitivity to production and installation tolerances, but the weak linear relationships between position and output signal (near the zero crossings) often lead to significant nonlinearity errors for calculating the angular displacement [11,12].

In the inductive position sensor field, the nonlinearity error is around one percent [7,13,14]. To reduce the nonlinearity error, the sensor structure needs to be optimized.We previously presented an inductive angle sensor optimized using response surface methodology [15]. For simplicity the original paper did not discuss the influence of the sensor stator on the nonlinearity errors. However, it is found that the stator affects the behavior of electromagnetic fields within its rotor, which plays a key role in the linearity of the inductive angle sensor.

Reducing the size of these transducers, often to less than a hundred micrometers, makes them less invasive to the environment in which they operate and improves their response speed. When information about the value of the quantity measured by the optical fiber sensor is transmitted as a change in phase, wavelength or spectrum, the sensor is immune to any mechanical or acoustical disturbance which may vary only with respect to the intensity of the transmitted signal [1].Recently, most optoelectronic research in the biomedical area has been focused on spectral measurement techniques [2�C4], Optical Coherence Tomography [5,6] and Raman scattering [7,8]. Optical fiber sensors have a considerable potential in biomedical application.

The use of low-coherence optical-fiber Fabry-Perot interferometric sensors [9�C12] as sensors of hematocrit levels of whole human blood offers advantages over classical analytical methods, such as: the lack of special preparation of whole human blood samples, the very small amount of blood needed for the measurements, and very short measurement times. In this article, the ability of a low-coherence optical fiber sensor using a Fabry-Perot interferometer with spectral signal processing to assess the hematocrit level in whole human blood without any special preparation under favorable conditions is presented. An optimal interferometer configuration giving a visibility of measured signals close to 1 has been achieved, and a series of measurements of the investigated liquids has been obtained by using this construction.2.

?Fiber-Optic Fabry-Perot InterferometerThe Fabry-Perot interferometer made from bulk optical components consists of two flat transparent plates P1 and P2, parallel to each other and separated by a distance d. Inner surfaces of P1 and P2 are coated with highly reflective layers L1 and L2. Expressions for intensity of light reflected from and transmitted by this interferometer are derived in most cases under three assumptions: (1) the interferometer is illuminated by a plane wave; (2) layers L1 and L2 have the same reflectivity; (3) the interferometer is lossless. Such a derivation yields well-known classic formulas [13]. In contrast, the fiber-optic Fabry-Perot interferometer, shown schematically in Figure 1, is illuminated by a divergent beam from a single-mode fiber.

Its reflective layers L1 and L2 may have different reflectivity and may not be parallel [10]. Moreover, Dacomitinib the layers and the medium between them may be absorbing. Consequently, the classic formulas do not hold.Figure 1.The construction of fiber-optic Fabry-Perot interferometer: L1, L2��the first and second reflective layers, respectively.In further discussing the fiber-optic Fabry-Perot interferometer, it will be assumed that layers L1 and L2, as well as the medium in the interferometer cavity, are non-absorbing.

For instance, a microfluidic ATP-bioluminescence sensor for the detection of airborne microbes using commercial available photo-diodes has been recently reported [27]. Although optical absorption detection is compatible with microfluidics, they suffer from relatively poor detection limits due to the short effective path length found in microfluidic channels [34]. Consequently, fluorescence detection remains the dominant optical detection technique in microfluidics. Here the conjugation of affinity markers (e.g. antibodies, DNA etc.) with fluorescent compounds like fluorescein isothiocyanate (FITC), phycoerythrin (PE) cyanin- or Alexa-dyes is most commonly used. Alternative approaches are based on the incorporation of two fluorescence molecules into the biosensor, using fluorescence resonance energy transfer (FRET) [35].

Other optical methods include chemiluminescence (CL), bioluminescence (BL) and Surface Plasmon Resonance (SPR) biosensors. While chemiluminescence describes the generation of light due to release of energy during a chemical reaction, SPR measures changes in refractive index caused by structural alterations in the vicinity of a thin film metal surface [36]. The numerous chemiluminescence (CL) applications in microfluidic analysis systems using immobilized enzymes, antibodies or nucleic acids have been recently described [37-39]. In turn, electroanalytical methods are highly compatible with micro- and nanomachining (MEMS) technology and can be segmented into current (amperometric), potential (potentiometric) or impedance (impediometric) techniques [40-43].

Evolving from ISFETs, a recent technology combines potentiometry and optical detection, known as light addressable potentiometric sensor (LAPS), that can be used for the detection of pathogen E. AV-951 coli [44]. Alternative detection methods for pathogen sensing include the application of silver dots for direct optical density measurements using a scanometric reader [45,46], or biosensors using resonance light scattering (RLS) techniques based on nanometer-sized metallic particles (mostly gold) covalently linked to antibodies. These metal colloidal particles radiate energy in the form of scattered light when illuminated by a white light source [47]. Altogether, LOC devices present themselves as a flexible technology platform that can be readily adapted to specific identification needs. A whole range of materials and mode of detection can be specifically selected for either low cost applications or high end analysis. Having reviewed the various materials and detection methods employed in lab-on-a-chip devices, we now provide a detail list of LOC studies grouped by class of target analytes.3.

Liu demonstrated that the accumulation of liquid water columns in the cathode flow channels reduces the effective electrochemical reaction area, limiting mass transfer and worsening cell performance [3]. Wang noted that liquid water management significantly affects PEMFC performance, especially at high current density [4]. Therefore, suitable water and thermal management should be used to ensure that the proton exchange membrane is sufficiently hydrated to maintain high proton conductivity.Li reviewed more than 100 references related to water management in proton exchange membrane fuel cells (PEMFCs), with a particular focus on water flooding, its diagnosis and mitigation [5].

Trabold applied neutron imaging to research the distribution of water flooding, detecting in situ variation in the amount of water that is produced in an operating fuel cell [6].

Tests that were performed by Zhang revealed that performance gradually worsened as relative humidity declined from 100% to 25% [7].Most investigations of voltage and humidity in PEMFCs involve the insertion of small sensors into the cells. For example, David examined the temperature distribution in fuel cells using Fiber Bragg grating technology. The result revealed a difference between the temperatures of the inlet and the outlet of 1 ��C [8]. Inman measured in-situ the reaction temperature in an operating fuel cell by placing five fiber temperature sensors in it [9].

Hinds employed commercial temperature and humidity sensors, with a large active area, in a single cell PEMFC [2].

Nishikawa cut the flow channel plate to install a commercial humidity sensor. This method yielded information about the interior, but the cost and assembly were problematic [10].Wang utilized Carfilzomib an infrared temperature device to measure external temperature distribution under various operating conditions [11]. Karimi observed the distribution of water within fuel cell GSK-3 stacks. His simulation results revealed that increasing the humidity promoted water flooding downstream [12]. Shimpalee simulated variations in temperature, humidity, and current in a PEMFC. His results demonstrated that water flooding downstream affected the fuel cell reaction, indirectly reducing the temperature and current [13].

In the aforementioned references, bipolar plates were cut and processed, and then sensors were inserted into fuel cells to measure internal physical values. This process can not only cause fuel leakage but also increase contact resistance. Along with invasive measurement, simulation can also identify water flooding. However, neither of these methods can be used to obtain accurate information on the interiors of fuel cells.