Using aspects for testing of embedded software: experiences from two industrial case studies

Aspect-oriented software testing is emerging as an important alternative to conventional procedural and object-oriented testing techniques. This paper reports experiences from two case studies where aspects were used for the testing of embedded software in the context of an industrial application. In the first study, we used code-level aspects for testing non-functional properties. The methodology we used for deriving test aspect code was based on translating high-level requirements into test objectives, which were then implemented using test aspects in AspectC++. In the second study, we used high-level visual scenario-based models for the test specification, test generation, and aspect-based test execution. To specify scenario-based tests, we used a UML2-compliant variant of live sequence charts. To automatically generate test code from the models, a modified version of the S2A Compiler, outputting AspectC++ code, was used. Finally, to examine the results of the tests, we used the Tracer, a prototype tool for model-based trace visualization and exploration. The results of the two case studies show that aspects offer benefits over conventional techniques in the context of testing embedded software; these benefits are discussed in detail. Finally, towards the end of the paper, we also discuss the lessons learned, including the technological and other barriers to the future successful use of aspects in the testing of embedded software in industry.     

Numerical optimization of conduction cooled current leads for low current applications

The basic design principles of current leads for superconducting magnets are well established but HTS materials and conduction cooled systems call for new numerical methods. In this paper the design of current leads was formulated as an optimization problem. Both time integration and finite differencing were examined as possible ways to compute the temperature distribution inside the leads. Three examples about optimization of conduction cooled as well as gas cooled systems are presented. First, the design of tubular normal conducting gas cooled current leads was studied. Second, normal conducting leads cooled with a two-stage cryocooler were examined. Third, the optimization was applied to current leads consisting of HTS tapes at the low temperature end of a normal conducting bar. The study took into account the magnetic field and temperature dependent voltage-current characteristics of the anisotropic Bi-2223 material. The results are compared with traditional analytical ones and the numerical optimization is shown to be an efficient design tool for both normal conducting and HTS current leads.     

Mannans as stabilizers of oil-in-water beverage emulsions

The stabilizing effect of spruce galactoglucomannan (GGM) on a model beverage emulsion system was studied and compared to that of guar gum and locust bean gum galactomannans, konjac glucomannan, and corn arabinoxylan. In addition, guar gum was enzymatically modified in order to examine the effect of the degree of polymerization and the degree of substitution of galactomannans on emulsion stability. Use of GGM increased the turbidity of emulsions both immediately after preparation and after storage of up to 14 days at room temperature. GGM emulsions had higher turbidity than the emulsions containing other mannans. The initial turbidity increased with increasing GGM content, but after 14 days storage at room temperature, the turbidity was the highest for GGM/oil ratio of 0.10:1 when ethanol-precipitated GGM was used. Increasing the storage temperature to +45 °C led to rapid emulsion breakdown, but a decrease in storage temperature increased emulsion stability after 14 days. Confocal microscopy showed that the average particle size in the bottom part of GGM emulsions stored for 14 days was smaller than 1 μm. A low degree of polymerization and a high degree of substitution of the modified galactomannans were associated with a decrease in emulsion turbidity.     

Titania and titania-silver nanoparticle deposits made by Liquid Flame Spray and their functionality as photocatalyst for organic- and biofilm removal

Titania and titania-silver nanoparticle deposits were made by Liquid Flame Spray technique, in which the liquid precursor is injected into a high temperature flame, where it will evaporate and nucleate to nanosize particles. One-step and two-step methods were used for preparation of titania-silver deposits. The amount of silver added was 1 wt%. The deposits were collected in the flame zone on steel and glass surfaces and were analyzed by TEM, EDS, XPS and SAXS. The titania deposits consisted of porous nanosized titania agglomerates of primary particles (~10 nm). With silver addition, small spherical silver metal particles (~2 nm) were detected on the agglomerates. An increase in the photocatalytic activity was verified by stearic acid decomposition and biofilm removal using Deinococcus geothermalis as the model organism.     

Detection of stressed oil palms from an airborne sensor using optimized spectral indices

Existing vegetation indices and red-edge techniques have been widely used for the assessment of vegetation status and vegetation health from remote-sensing instruments. This study proposed and applied optimized Airborne Imaging Spectrometer for Applications (AISA) airborne hyperspectral indices in assessing and mapping stressed oil palm trees. Six vegetation indices, four red-edge techniques, a standard supervised classifier and three optimized AISA spectral indices were compared in mapping diseased oil palms using AISA airborne hyperspectral imagery. The optimized AISA spectral indices algorithms used newly defined reflectance values at wavelength locations of 734 nm (near-infrared (NIR)) and 616 nm (red). The selection of these two bands was based on laboratory statistical analysis using field spectroradiometer reflectance data. These two bands were then applied to the AISA airborne hyperspectral imagery using the three optimized algorithms for AISA data. The newly formulated AISA hyperspectral indices were D2 = R ₆₁₆/R ₇₃₄, normalized difference vegetation index a (NDVIa)?=?(R ₇₃₄ – R ₆₁₆)/(R ₇₃₄?+?R ₆₁₆) and transformed vegetation index a (TVIa)?=?((NDVIa?+?0.5)/(abs (NDVIa?+?0.5))?×?[abs (NDVIa?+?0.5)]^1/2. The classification results from the optimized AISA hyperspectral indices were compared with the other techniques and the optimized AISA spectral indices obtained the highest overall accuracy. D2 and NDVIa obtained 86% of overall accuracy followed by TVIa with 84% of overall accuracy.     

Flexibility as a design driver [systems analysis]

Ambiguous requirements and those that emerge late in the design cycle often complicate development and throw off established schedules. The authors describe how reformulating information missing at the design stage into a flexibility requirement can turn the absent data into a design driver. Designers constantly run into issues that have yet to be understood through specification, system design, or standardization. Changes in evolving technologies and businesses often result in unstable system requirements. Reliable hardware or mechanical details might not be available until very late in the development process. Yet engineers must initiate software development even though some subsystem details have not yet been completely defined. Missing information and related flexibility requirements can lead to a design plagued by many well-known problems that affect performance, modularity, scalability, and clear separation of concerns. In some cases, developers must sacrifice rules of thumb to maintain planned development schedules. Providing flexibility for everything in the system isn't possible, so developers will always need to determine the static requirements and explicitly state where continued development and rapid modification require flexibility     

On-Line Characterization of Morphology and Water Adsorption on Fumed Silica Nanoparticles

The first wetting layer on solid nanoparticles has direct implications on the roles these particles play in industrial processes and technological applications as well as in the atmosphere. We present a technique for online measurements of the adsorption of the first few water layers onto insoluble aerosol nanoparticles. Atomized fumed silica nanoparticles were dispersed from aqueous suspension and their hygroscopic growth factors (HGF) and number of the adsorbed water layers at subsaturated conditions were measured using a nanometer hygroscopic tandem differential mobility analyzer (HTDMA). Particle morphology was characterized by electron microscopy and particle density was determined by mobility analysis. The HGFs of the size-selected particles at mobility diameters from 10 to 50 nm at 90% relative humidity (RH) varied from 1.05 to 1.24, corresponding to 2–6 layers of adsorbed water. The morphology of the generated fumed silica nanoparticles varied from spheres at 8–10 nm to agglomerates at larger diameters with effective density from 1.7 to 0.8 g/cm³ and fractal dimension of 2.6. The smallest spheres and agglomerates had the highest HGFs. The smallest particles with diameters of 8 and 10 nm adsorbed two to three water layers in subsaturated conditions, which agreed well with the Frenkel, Halsey, and Hill (FHH) isotherm fitting. In comparison to the small spheres or large agglomerates, the compact agglomerate structure containing a few primary particles increased the number of adsorbed water layers by a factor of ～1.5. This was probably caused by the capillary effect on the small cavities between the primary particles in the agglomerate.