Learning Paths to Offshore Outsourcing: From Cost Reduction to Knowledge Seeking

A corporation's offshore outsourcing may be seen as the result of a discrete, strategic decision taken in response to an increasing pressure from worldwide competition. However, empirical evidence of a representative cross-sector sample of international Danish firms indicates that offshore sourcing in low-cost countries is best described as a learning-by-doing process in which the offshore outsourcing of a corporation goes through a sequence of stages towards sourcing for innovation. Initially, a corporation's outsourcing is driven by a desire for cost minimization. Over a period of time the outsourcing experience lessens the cognitive limitations of decision-makers as to the advantages that can be achieved through outsourcing in low-cost countries: the insourcer/vendor may not only offer cost advantages, but also quality improvement and innovation. The quality improvements that offshore outsourcing may bring about evoke a realization in the corporation that even innovative processes can be outsourced.     

监控线路连接交流负载的电压定时器

一个简单的监控线路连接的交流负载供电所用的时间(如图1).可以转换成标准单线协议所用时间的计数值.当给交流负载供电时,光隔离器以交流线路频率向单线计数器DS2423 IC的输入端提供脉冲.     

A unifying framework for automatic and semi-automatic segmentation of vertebrae from radiographs using sample-driven active shape models

Segmentation of vertebral contours is an essential task in the design of imaging biomarkers for osteoporosis based on vertebra shape or texture. In this paper, we propose a novel automatic segmentation technique which can optionally be constrained by the user. The proposed technique solves the segmentation problem in a hierarchical manner. In the first phase, a coarse estimate of the overall spine alignment and the vertebra locations is computed using a sampling scheme. These samples are used to initialize a second phase of active shape model search, under a nonlinear model of vertebra appearance. The search is constrained by a conditional shape model, based on the variability of the coarse spine location estimates. In supplement, we describe an approach for manual initialization of the segmentation procedure as a simple set of constraints on the fully automatic technique. The technique is evaluated on a data base of 157 manually annotated lumbar radiographs, resulting in a final mean point-to-contour error of $0.81~\pm ~0.98$  mm for automatic segmentation. The results outperform the previous work in automatic vertebra segmentation in terms of both segmentation accuracy and failure rate, offering a both automatic and semi-automatic approach in one unifying framework.     

A method for simultaneous strong stabilization of linear time-varying systems

This paper considers the problem of simultaneously stabilizing a finite collection of linear time-varying systems via a stable linear time-varying controller. The class of controllers proposed in this paper are linear, time varying and infinite dimensional. However, the computer implementation of these controllers would not be computationally more demanding than the computer implementation of a finite-dimensional linear time-varying controller.     

Robust incoherent control of qubit systems via switching and optimisation

A robust incoherent quantum control scheme via projective measurements plus unitary transformations is proposed for driving a qubit system from an unknown initial mixed state to an arbitrary target pure state. This scheme consists of two main steps: projective measurement on the initial mixed state and optimal control between two pure states. The first step projects the initial state into an eigenstate of the qubit system by projective measurement and guarantees that the proposed scheme is robust to different initial mixed states. The second step finds a set of suitable optimal controls to drive the qubit system from the conditional eigenstate to the target pure state. The connection between the two steps is accomplished by a switching strategy. To accomplish the second step, two approaches are presented in detail. These approaches are time-optimal transition with unbounded control and bang-bang control with minimal switches. The minimal time and minimal number of switches in these approaches can be calculated by simple analytical expressions. The proposed approaches provide two relatively straightforward optimal design methods.     

A decision support framework for metrics selection in goal-based measurement programs: GQM-DSFMS

Software organizations face challenges in managing and sustaining their measurement programs over time. The complexity of measurement programs increase with exploding number of goals and metrics to collect. At the same time, organizations usually have limited budget and resources for metrics collection. It has been recognized for quite a while that there is the need for prioritizing goals, which then ought to drive the selection of metrics. On the other hand, the dynamic nature of the organizations requires measurement programs to adapt to the changes in the stakeholders, their goals, information needs and priorities. Therefore, it is crucial for organizations to use structured approaches that provide transparency, traceability and guidance in choosing an optimum set of metrics that would address the highest priority information needs considering limited resources. This paper proposes a decision support framework for metrics selection (DSFMS) which is built upon the widely used Goal Question Metric (GQM) approach. The core of the framework includes an iterative goal-based metrics selection process incorporating decision making mechanisms in metrics selection, a pre-defined Attributes/Metrics Repository, and a Traceability Model among GQM elements. We also discuss alternative prioritization and optimization techniques for organizations to tailor the framework according to their needs. The evaluation of the GQM-DSFMS framework was done through a case study in a CMMI Level 3 software company.     

Robust decentralized stabilization of Markovian jump large-scale systems: A neighboring mode dependent control approach

This paper is concerned with the decentralized stabilization problem for a class of uncertain large-scale systems with Markovian jump parameters. The controllers use local subsystem states and neighboring mode information to generate local control inputs. A sufficient condition involving rank constrained linear matrix inequalities is proposed for the design of such controllers. A numerical example is given to illustrate the developed theory.     

Strength variability of single flax fibres

Due to the typical large variability in the measured mechanical properties of flax fibres, they are often employed only in low grade composite applications. The present study aims to investigate the reasons for the variability in tensile properties of flax fibres. It is found that an inaccuracy in the determination of the cross-sectional area of the fibres is one major reason for the variability in properties. By applying a typical circular fibre area assumption, a considerable error is introduced into the calculated mechanical properties. Experimental data, together with a simple analytical model, are presented to show that the error is increased when the aspect ratio of the fibre cross-sectional shape is increased. A variability in properties due to the flax fibres themselves is found to originate from the distribution of defects along the fibres. Two distinctive types of stress–strain behaviours (linear and nonlinear) of the fibres are found to be correlated with the amount of defects. The linear stress–strain curves tend to show a higher tensile strength, a higher Young’s modulus, and a lower strain to failure than the nonlinear curves. Finally, the fibres are found to fracture by a complex microscale failure mechanism. Large fracture zones are governed by both surface and internal defects; and these cause cracks to propagate in the transverse and longitudinal directions.     

Improved Bounds for Functions Related to Busy Beavers

Consider Turing machines that use a tape infinite in both directions, with the tape alphabet {0,1} . Rado's busy beaver function, ones(n), is the maximum number of 1's such a machine, with n states, started on a blank (all-zero) tape, may leave on its tape when it halts. The function ones(n) is non-computable; in fact, it grows faster than any computable function. Other functions with a similar nature can be defined also. All involve machines of n states, started on a blank tape. The function time(n) is the maximum number of moves such a machine may make before halting. The function num(n) is the largest number of 1's such a machine may leave on its tape in the form of a single run; and the function space(n) is the maximum number of tape squares such a machine may scan before it halts. This paper establishes new bounds on these functions in terms of each other. Specifically, we bound time(n) by num(n+o(n)), improving on the previously known bound num(3n+6) . This result is obtained using a kind of ``self-interpreting' Turing machine. We also improve on the trivial relation space(n) ≤ time(n) , using a technique of counting crossing sequences. Received July 18, 2000, and in revised form October 10, 2001. Online publication February 20, 2002.