Miguel Dias Pereira

NDT on Aluminum Aircraft Plates based on Eddy Current Sensing and Image Processing

Non-destructive testing (NDT) applied to aluminum aircraft plates refers to testing the plates without impairing its worth or functional properties. In the work a NDT system architecture based on the use of fully inductive probes with excitation and sensing coils and hybrid probe with an excitation coil and a magneto-resistance as sensing element is presented. Appropriate signal conditioning and signal acquisition modules are included in the system. Using an X-Y plotter the probe is moved over the aluminum plate under test and a set of eddy current images is obtained. Image conversions, image filtering, image segmentation with feature extraction and geometrical characteristics of the detected cracks determines the capability of the designed and implemented software for non destructive testing. A practical approach concerning the optimal filtering of the image for a better segmentation and feature extraction was included in the paper. Several results concerning the aluminum plate crack detection and geometrical characterization are included.

Improving celerity of heavy metals measurements

This paper presents a low cost measurement system to identify heavy metals dissolved in water and to evaluate their concentrations. Two main objectives of the proposed measurement system must be underlined, both related with measurement celerity. The first one improves measurement celerity by reducing the requested time to remove dissolved oxygen contained in the water sample. The second one proposes a flexible voltammetry voltage scan that improves measurement celerity using a fast voltage scan that identifies the heavy metals dissolved in the solution with higher concentrations and then a slower voltage scan is performed around the voltammogram current peaks associated with the heavy metals with higher concentrations.

Pervasive sensing and computing for wheelchairs users health assessment

Optimal cardiovascular, respiratory and motor activity assessment of wheelchair users using modern technologies of pervasive sensing and computing represents the main objective of the running project. In order to extract accurate information about physiological parameters provided by physical measurement channel and also to estimate additional parameters associated with virtual measurement channels advanced processing algorithms are implemented on embedded computing platform of the wheelchair. Will be implemented ballistocardiography, electrocardiography, plethysmography, skin conductance smart sensing devices that extract the health status information in unobtrusive way. The informations from the smart sensors, as from the localization units based on GPS technology (for outdoor localization) and RFID technology (for indoor localization) are processed on the embedded computing platform based on DSP that also assure the wireless data transmission to the human machine interface (HMI) expressed by a tablet that run mobile OS such as Android or Windows Mobile. A web based information system is used as part of the electronic health record that assure the remote health care assistance of the wheelchair users.

Embedding IEEE 1451.4 smart sensing nodes in a Wireless Air Quality Monitoring Network

An hybrid smart sensor network that includes a set of sensing nodes designed and implemented according with the specifications of the Smart Transducer Interface Standard, IEEE 1451, is presented. The communication between the nodes and the host computer that is used for advanced air quality data processing and data storage is performed through a wireless connection that was developed using a low cost, high performance radio device. Regarding the IEEE 1451 implementation the nodes put together a minimal Dot4 Network Capable Application Processor (NCAP) with a data acquisition unit that reads the analog signals from an IEEE 1451.4 Transducer Unit with air quality, temperature and relative humidity sensors using a Mixed-Mode Interface (MMI). The LabVIEW program implemented on the computer performs the automatic identification of the node analog sensors through TEDS reading, air quality data correction using the temperature and relative humidity information, data logging and pollution alarm generation. The use of Virtual TEDS solution for the sensing nodes is also considered.

Colored light-to-voltage converters based absorbance meter

A low-cost absorbance meter for dissolved compounds identification and concentration evaluation is proposed. The main elements of the system are a white light source and a set of three colored light-to-voltage converters. The system includes auto-calibration capabilities and compensates spectral sensitivity variation of the optical source and detectors. A neural network-processing scheme is used to identify dissolved compounds and evaluate the correspondent concentrations.

A Smart Measurement and Stimulation System to Analyze and Promote Non-Nutritive Sucking of Premature Babies

A Smart Measurement and Stimulation System to Analyze and Promote Non-Nutritive Sucking of Premature Babies This paper presents a complete system that can be used to measure and to stimulate sucking abilities of premature babies. The system integrates measurement and stimulation capabilities that can be used separately or together. The sensing unit includes mainly a pressure sensor, a conditioning circuit and a microcontroller device. The stimulation unit includes a miniature and low-cost pneumatic pump, an electro-valve, and a conditioning circuit. The microcontroller is shared by both units. The systems software includes a routine that is used to define the waveform parameters of the synthesized pressure signals, a wavelet based routine that is used to process measurement data and a software component to perform fault detection and diagnosis. The system capabilities include self-calibration and self-testing. Throughout the paper several implementation details and experimental results will be presented and discussed.

Advanced signal processing techniques to measure and classify non-nutritive suction of premature and newly born babies

This paper addresses a crucial problem that exists frequently in data processing of biomedical signals. Almost all of these signals have a very low signal-to-noise ratio and a careful hardware and software development are essential to obtain a meaningful and accurate measurement. This is really important, especially when human diagnostics are at stage, since errors can affect patients health and sometimes endanger life. This paper proposes a measurement solution to obtain an accurate measurement of non-nutritive suctions (NSS) in premature babies in order to avoid excessive costs caused by superfluous occupation of intensive care units. Proposed signal processing techniques, based on Gaussian and progressive polynomial curve fitting techniques, are used to obtain an accurate value of sucking parameters. The application domain of the proposed measurement system can be extended to other biomedical measurement systems whenever low pressure measurements are required, particularly those related with nutritive sucking (NS).

Using neural network techniques in environmental sensing and measurement systems to compensate for the effects of influence quantities

Multiple influence quantities affect environmental sensing and measurement (ESM) systems. Accounting for their variations over time promotes metrological comparability and traceability of measurement results. Using suitable data processing techniques allows identification of the main influence quantities that affect the measurement result of a given quantity and allows evaluation of the associated compensation coefficients. This paper presents several concerns related to implementation of ESM systems, paying particular attention to the impact of the multiple of measuring and influence quantities on system performance. A review of different data processing techniques to compensate for the effects of influence quantities is presented. A case study based on water conductivity measurements is used to illustrate the capability of artificial neural network (ANN) based techniques and to cope with errors due to a low number of measurement values and due to collinear effects between influence quantities.

Service-oriented distributed measurement and control systems: A case study

The paper presents a Distributed Measurement and Control System (DMCS) applied to a physical process equipped with all the instrumentation needed to run control loops for temperature, level, flow and pressure, quantities widely found in the process industry. The system is supported by a service-oriented middleware platform based on the IEEE 1451.1 std and the Windows Communication Foundation (WCF). Tests were made to evaluate the performance of the system in terms of sampling frequency and behavior of control loops. The methodology of each experiment is described, results are analyzed and conclusions are extracted.

Performance evaluation of a Web-Service-based DMCS

The paper describes a set of experiments conducted on a service-oriented middleware infrastructure in order to evaluate its performance and applicability in the context of Distributed Measurement and Control Systems (DMCS). The infrastructure, entirely based on Web Services, was built using the Windows Communication Foundation (WCF), a software package released by Microsoft to develop distributed applications. The experiments were performed on a real plant equipped with all the instrumentation needed to run control loops for pressure, level, flow and temperature, quantities widely found in the process industry. The work focus on measuring the time delays associated with control loops and remote calls. The methodology of each experiment is described, results are presented and conclusions are drawn.