Adaptive polygonization of geometrically constrained surfaces

n -dimensional space, where n>3. This definition can be used for given surfaces that are implicit or parametric. This paper presents a robust, adaptive polygonization algorithm for evaluating and visualizing geometrically constrained surfaces. Let be the constrained surface, a 2-surface in n-space, and let π() be its projection into the subspace spanned by the first three coordinates. Our polygonization algorithm computes π(). The method works directly with the n-space representation, but performs all major computations in 3-space. Techniques for triangulation, polygon decimation, and local refinement are also presented.     

Using regression splines to assess the impact of information technology investments on productivity in the health care industry

Abstract. This paper explores the impact of information technology (IT) investments on productivity using a new technique, multivariate adaptive regression splines (MARS). We believe that it provides additional insights on the nature of the impact of IT investments on productivity. The results from our study are compared with findings from a previous study that has also used the same data set. While the results of a previous study indicate that IT investments have a positive but uniform impact on productivity, our study suggests that the impact of IT on productivity is not uniform but is contingent on other complementary factors. Our findings describe that the complementary relationship exists between IT and non‐IT related investments. Thus, improved organizational productivity cannot be expected from investment in IT alone but only together with non‐IT investments. Our findings also point out that further investment may not necessarily bring on higher organizational productivity.     

High-temperature single-crystal 3C-SiC capacitive pressure sensor

Single-crystal 3C-silicon carbide (SiC) capacitive pressure sensors are proposed for high-temperature sensing applications. The prototype device consists of an edge-clamped circular 3C-SiC diaphragm with a radius of 400 /spl mu/m and a thickness of 0.5 /spl mu/m suspended over a 2-/spl mu/m sealed cavity on a silicon substrate. The 3C-SiC film is grown epitaxially on a 100-mm diameter <100> silicon substrate by atmospheric pressure chemical vapor deposition. The fabricated sensor demonstrates a high-temperature sensing capability up to 400/spl deg/C, limited by the test setup. At 400/spl deg/C, the device achieves a linear characteristic response between 1100 and 1760 torr with a sensitivity of 7.7 fF/torr, a linearity of 2.1%, and a hysterisis of 3.7% with a sensing repeatability of 39 torr (52 mbar). A wide range of sensor specifications, such as linear ranges, sensitivities, and capacitance values, can be achieved by choosing the proper device geometrical parameters.     

Nitride-based high power flip-chip near-UV leds with reflective submount

Nitride-based high power flip-chip near-ultraviolet (UV) light emitting diodes (LEDs) with a reflective mirror are fabricated by depositing Al onto a Si submount. It is demonstrated that the Al layer coated onto a Si submount can effectively reflect downward emitting photons for flip-chip LEDs. Although the operation voltage of the proposed LEDs is slightly increased, it is found that the output power is at least 30% higher than that of conventional LEDs. It is also found that flip-chip near-UV LEDs are more reliable than conventional non-flip-chip LEDs     

Effect of overload on fatigue crack retardation of aerospace Al-alloy laser welds using crack-tip plasticity analysis

Investigations on fatigue crack growth retardation due to single tensile and periodic multiple over load in strength undermatched laser beam welded 3.2 mm thick aerospace grade aluminium alloy 2139-T8 sheets are conducted. The effect of overload on the fatigue crack propagation behaviours of the homogenous base metal and welded panels (200 mm wide, centre cracked) was compared using experimental and FE analysis methods. The effective crack tip plasticity has been determined in homogeneous M(T) specimens using Irwin’s method and in both homogeneous and laser welded specimen by calculating crack tip plastic strain using FE analysis for single tensile overload. The crack retardation due to the overload in welded specimens is described by the Wheeler Model. The crack tip plastic zone size in the welded specimen was determined by FE analysis using maximum plastic zone extension at the mid sheet thickness. The results show that the Wheeler Model can be implemented to the highly heterogeneous undermatched weld to describe the crack retardation in fatigue following single tensile overload. Fatigue crack growth retardation due to single overload is found to be larger than the base metal. However, after periodic multiple overload, shorter crack retardation has occurred for undermatched welds than the base metal.     

Efficient location and allocation strategies for undesirable facilities considering their fundamental properties

The phrase “not in my backyard” (NIMBY) refers to the well-known social phenomena in which residents oppose the construction or location of undesirable facilities near their homes. Examples of such facilities include electric transmission lines, recycling centers and crematoria. Due to the opposition typically encountered in constructing an undesirable facility, the facility planner should understand the nature of the NIMBY phenomena and consider it as a key factor in determining facility location. We examine the characteristics of NIMBY phenomena and suggest two alternative mathematical optimization models with the objective of minimizing the total degree of NIMBY sentiments. Genetic algorithms are proposed to solve our linear and nonlinear integer programs. The results obtained via genetic algorithms for our linear integer programs are compared with those of CPLEX to evaluate their performance. The nonlinear programs are tested with various allocation policies. Sensitivity analysis is conducted about several system parameters.     

Spatiotemporal bag-of-features for early wildfire smoke detection

Wildfire smoke detection is particularly important for early warning systems, because smoke usually rises before flames arise. Therefore, this paper presents an automatic wildfire smoke detection method using computer vision and pattern recognition techniques. First, candidate blocks are identified using key-frame differences and nonparametric smoke color models to detect smoke-colored moving objects. Subsequently, three-dimensional spatiotemporal volumes are built by combining the candidate blocks in the current key-frame with the corresponding blocks in previous frames. A histogram of oriented gradient (HOG) is extracted, and a histogram of oriented optical flow (HOOF) is extracted as a temporal feature based on the fact that the direction of smoke diffusion is upward owing to thermal convection. From spatiotemporal features of training data, a visual codebook and a bag-of-features (BoF) histogram are generated using our proposed weighting scheme. For smoke verification, a random forest classifier is built during the training phase using the BoF histogram. The random forest with the BoF histogram can increase the detection accuracy performance when compared with related methods and allow smoke detection to be carried out in near real time.     

Pixel‐Structured Scintillator with Polymeric Microstructures for X‐Ray Image Sensors

We introduce a pixel‐structured scintillator realized on a flexible polymeric substrate and demonstrate its feasibility as an X‐ray converter when it is coupled to photosensitive elements. The sample was prepared by filling Gd₂O₂S:Tb scintillation material into a square‐pore‐shape cavity array fabricated with polyethylene. For comparison, a sample with the conventional continuous geometry was also prepared. Although the pixelated geometry showed X‐ray sensitivity of about 58% compared with the conventional geometry, the resolving power was improved by about 70% above a spatial frequency of 3 mm^?1. The spatial frequency at 10% of the modulation‐transfer function was about 6 mm^?1.     

Design and implementation of multicasting for multi-channel multi-interface wireless mesh networks

Multicasting is a useful communication method in wireless mesh networks (WMNs). Many applications in WMNs require efficient and reliable multicast communications, i.e., high delivery ratio with low overhead among a group of recipients. In spite of its significance, little work has been done on providing such multicast service in multi-channel WMNs. Traditional multicast protocols for wireless and multi-hop networks tend to assume that all nodes, each of which is equipped with a single interface, collaborate on the same channel. This single-channel assumption is not always true, as WMNs often provide nodes with multiple interfaces to enhance performance. In multi-channel and multi-interface (MCMI) WMNs, the same multicast data must be sent multiple times by a sender node if its neighboring nodes operate on different channels. In this paper, we try to tackle the challenging issue of how to design a multicast protocol more suitable for MCMI WMNs. Our multicast protocol builds multicast paths while inviting multicast members, and tries to allocate the same channel to neighboring members in a bottom-up manner. By unifying fixed channels of one-hop multicast neighbors, the proposed algorithm can improve the performance such as reducing multicast data transmission overhead and delay, while managing a successful delivery ratio. In order to prove such expectation on the performance, we have implemented and evaluated the proposed solution on the real testbed having the maximum 24 nodes, each of which is equipped with two IEEE 802.11a Atheros WLAN cards.