Run-to-run failure detection and diagnosis using neural networks and Dempster-Shafer theory: an application to excimer laser ablation

The formation of microvias in multilayer substrates is a critical factor in microelectronic packaging manufacturing. Such microstructures can be produced efficiently by excimer laser ablation. Thus, laser ablation systems are evolving to a level where the need to offset high capital equipment investment and lower equipment downtime are imminent. This paper presents a methodology for inline failure detection and diagnosis of the excimer laser ablation process. The methodology employs response data originating directly from the equipment and characterization of microvias formed by the ablation process. Neural network (NN) models are trained and validated based on this data to generate evidential belief for potential sources of deviations in the responses. Dempster-Shafer (D-S) theory is adopted for evidential reasoning. Successful failure detection is achieved in 100% of 19 possible failure scenarios. Moreover, successful failure diagnosis is also achieved with only a single false alarm occurring in the 19 failure scenarios.     

Managing obesity through mobile phone applications: a state-of-the-art review from a user-centred design perspective

Evidence has shown that the trend of increasing obesity rates has continued in the last decade. Mobile phone applications, benefiting from their ubiquity, have been increasingly used to address this issue. In order to increase the applications’ acceptance and success, a design and development process that focuses on users, such as user-centred design, is necessary. This paper reviews reported studies that concern the design and development of mobile phone applications to prevent obesity, and analyses them from a user-centred design perspective. Based on the review results, strengths and weaknesses of the existing studies were identified. Identified strengths included: evidence of the inclusion of multidisciplinary skills and perspectives; user involvement in studies; and the adoption of iterative design practices. Weaknesses included the lack of specificity in the selection of end-users and inconsistent evaluation protocols. The review was concluded by outlining issues and research areas that need to be addressed in the future, including: greater understanding of the effectiveness of sharing data between peers, privacy, and guidelines for designing for behavioural change through mobile phone applications.     

Bacteremia in a community hospital: spectrum and mortality

The problem of bacteremia was studied at our 500-bed community hospital during 1974-1975. All patients with positive blood cultures for clinically significant organisms were analyzed with respect to age, sex, host factors, type, dose, and time of institution of therapy with antibiotics, and mortality. Of 142 episodes of bacteremia, outcome according to underlying host factors showed that 12/15 died in the rapidly, fatal group, 12/24 died in the ultimately fatal group, and in the nonfatal group 25/103 died. Mortality for Gram-positive bacteremia was 37% (7/54) for Gram-negative bacteremia in the nonfatal group. In addition to underlying host factors, inappropriate antibiotic therapy also contributed to higher mortality in patients with gram-negative bacteremia.     

Design and implementation of a parallel solution to the cutting stock problem

The constrained 2D cutting stock problem is an irregular problem with dynamic data structures, highly variable amounts of computation per task, and unpredictable amounts and patterns of communication. This paper describes the design and implementation of a parallel solution to this problem on a cluster of workstations and a distributed memory multicomputer. The key element of our parallel solution is the replication of an important data structure on all processors. By exploiting properties of the cutting stock problem which allow the use of relaxed consistency mechanisms, our approach is able to reduce the overheads for communication and synchronization in comparison to approaches that partition the data structure among processors. A token-based lazy release consistency protocol is used to ensure mutual exclusion and maintain consistency, and a randomized work-stealing protocol is employed to dynamically balance work among processors. Good speedups are reported for three benchmark problems executed on two distributed memory platforms: a cluster of workstations interconnected by a 10 Mbit/s Ethernet and an Intel Paragon. © 1998 John Wiley & Sons, Ltd.     

A New Fourth-Order Compact Off-Step Discretization for the System of 2D Nonlinear Elliptic Partial Differential Equations

This paper discusses a new fourth-order compact off-step discretization for the solution of a system of two-dimensional nonlinear elliptic partial differential equations subject to Dirichlet boundary conditions. New methods to obtain the fourth-order accurate numerical solution of the first order normal derivatives of the solution are also derived. In all cases, we use only nine grid points to compute the solution. The proposed methods are directly applicable to singular problems and problems in polar coordinates, which is a main attraction. The convergence analysis of the derived method is discussed in detail. Several physical problems are solved to demonstrate the usefulness of the proposed methods.     

Supply chain information technologies and organisational initiatives: complementary versus independent effects on agility and firm performance

This research investigates the roles of supply chain information technologies (SCIT) and supply chain organisational initiatives (SCOI) in engendering agility and business performance in manufacturing firms. We examine two competing models, both of which incorporate agility as a mediator between the use of SCIT or SCOI and firm performance; the models differ in how the impacts of SCIT and SCOI are manifest. In one model, SCIT and SCOI are hypothesised to have separate effects on agility, which then impacts firm performance; in the second model, complementarities, or the interaction of SCIT and SCOI, impacts agility directly. Structural equation modelling results show that agility is full mediator, related to firm performance in both models. Further, the model with complementary interactions fits better. These results have implications for how manufacturing firms can position their investments in SCIT and SCOI to enhance agility and overall performance.     

Colloid‐in‐Liquid Crystal Gels that Respond to Biomolecular Interactions

This paper advances the design of stimuli‐responsive materials based on colloidal particles dispersed in liquid crystals (LCs). Specifically, thin films of colloid‐in‐liquid crystal (CLC) gels undergo easily visualized ordering transitions in response to reversible and irreversible (enzymatic) biomolecular interactions occurring at the aqueous interfaces of the gels. In particular, LC ordering transitions can propagate across the entire thickness of the gels. However, confinement of the LC to small domains with lateral sizes of ～10 μm does change the nature of the anchoring transitions, as compared to films of pure LC, due to the effects of confinement on the elastic energy stored in the LC. The effects of confinement are also observed to cause the response of individual domains of the LC within the CLC gel to vary significantly from one to another, indicating that manipulation of LC domain size and shape can provide the basis of a general and facile method to tune the response of these LC‐based physical gels to interfacial phenomena. Overall, the results presented in this paper establish that CLC gels offer a promising approach to the preparation of self‐supporting, LC‐based stimuli‐responsive materials.     

Simulation of salinity effects on past, present, and future soil organic carbon stocks

Soil organic carbon (SOC) models are used to predict changes in SOC stocks and carbon dioxide (CO(2)) emissions from soils, and have been successfully validated for non-saline soils. However, SOC models have not been developed to simulate SOC turnover in saline soils. Due to the large extent of salt-affected areas in the world, it is important to correctly predict SOC dynamics in salt-affected soils. To close this knowledge gap, we modified the Rothamsted Carbon Model (RothC) to simulate SOC turnover in salt-affected soils, using data from non-salt-affected and salt-affected soils in two agricultural regions in India (120 soils) and in Australia (160 soils). Recently we developed a decomposition rate modifier based on an incubation study of a subset of these soils. In the present study, we introduce a new method to estimate the past losses of SOC due to salinity and show how salinity affects future SOC stocks on a regional scale. Because salinity decreases decomposition rates, simulations using the decomposition rate modifier for salinity suggest an accumulation of SOC. However, if the plant inputs are also adjusted to reflect reduced plant growth under saline conditions, the simulations show a significant loss of soil carbon in the past due to salinization, with a higher average loss of SOC in Australian soils (55 t C ha(-1)) than in Indian soils (31 t C ha(-1)). There was a significant negative correlation (p < 0.05) between SOC loss and osmotic potential. Simulations of future SOC stocks with the decomposition rate modifier and the plant input modifier indicate a greater decrease in SOC in saline than in non-saline soils under future climate. The simulations of past losses of SOC due to salinity were repeated using either measured charcoal-C or the inert organic matter predicted by the Falloon et al. equation to determine how much deviation from the Falloon et al. equation affects the amount of plant inputs generated by the model for the soils used in this study. Both sets of results suggest that saline soils have lost carbon and will continue to lose carbon under future climate. This demonstrates the importance of both reduced decomposition and reduced plant input in simulations of future changes in SOC stocks in saline soils.     

Effect of micro-powder suspension and ultrasonic vibration of dielectric fluid in micro-EDM processes—Taguchi approach

Solutions are needed for increasing the material removal rate without degrading surface quality in micro-electrical discharge machining (μ-EDM). This paper presents a new method that consists of suspending micro-MoS₂ powder in dielectric fluid and using ultrasonic vibration during μ-EDM processes. The Taguchi method is adopted to ascertain the optimal process parameters to increase the material removal rate of dielectric fluid containing micro-powder in μ-EDM using a L₁₈ orthogonal array. Pareto analysis of variance is employed to analyze the four machining process parameters: ultrasonic vibration of the dielectric fluid, concentration of micro-powder, tool electrode materials, and workpiece materials. The results show that the introduction of MoS₂ micro-powder in dielectric fluid and using ultrasonic vibration significantly increase the material removal rate and improves surface quality by providing a flat surface free of black carbon spots.     

A fault-tolerant algorithm for replicated data management

We examine the tradeoff between message overhead and data availability that arises in the design of fault-tolerant algorithms for replicated data management in distributed systems. We propose a property called asymptotically high resiliency which is useful for evaluating the fault-tolerance of replica control algorithms and distributed mutual exclusion algorithms. We present a new algorithm for replica control that can be tailored (through a design parameter) to achieve the desired balance between low message overhead and high data availability. Further, we show that for a message overhead of O(√(Nlog N)), our algorithm can achieve asymptotically high resiliency