Exploring an effective model of new product development in medical devices: A knowledge cluster approach

This study proposes a new product development (NPD) model that aims to improve the effectiveness of innovative NPD in the medical devices. By adopting open innovation theory and applying an in-depth investigation methodology, this paper proposes a knowledge cluster that improves the integration of interdisciplinary human resources and enhances the acquirement of innovative technologies. A knowledge cluster approach helps gather, organise, synthesise, and accumulate knowledge in order to become the impetus for innovation. Although enterprises are no longer the principals of research and development, they should still be capable of integrating professional physicians, external groups, and individuals through the knowledge cluster platform. However, in order to support an effective NPD model, enterprises should provide adequate incentives and trust to external individuals or groups willing to contribute their expertise and knowledge to this knowledge cluster platform. Copyright © 2013 Inderscience Enterprises Ltd.

Student-t processes as alternatives to Gaussian processes

We investigate the Student-t process as an alternative to the Gaussian process as a non-parametric prior over functions. We derive closed form expressions for the marginal likelihood and predictive distribution of a Student-t process, by integrating away an inverse Wishart process prior over the co-variance kernel of a Gaussian process model. We show surprising equivalences between different hierarchical Gaussian process models leading to Student-t processes, and derive a new sampling scheme for the inverse Wishart process, which helps elucidate these equivalences. Overall, we show that a Student-t process can retain the attractive properties of a Gaussian process - a nonparamet-ric representation, analytic marginal and predictive distributions, and easy model selection through covariance kernels - but has enhanced flexibility, and predictive covariances that, unlike a Gaussian process, explicitly depend on the values of training observations. We verify empirically that a Student-t process is especially useful in situations where there are changes in covariance structure, or in applications such as Bayesian optimization, where accurate predictive covariances are critical for good performance. These advantages come at no additional computational cost over Gaussian processes.

A global model for effective use and evaluation of e-learning in health.

Healthcare systems worldwide face a wide range of challenges, including demographic change, rising drug and medical technology costs, and persistent and widening health inequalities both within and between countries. Simultaneously, issues such as professional silos, static medical curricula, and perceptions of "information overload" have made it difficult for medical training and continued professional development (CPD) to adapt to the changing needs of healthcare professionals in increasingly patient-centered, collaborative, and/or remote delivery contexts. In response to these challenges, increasing numbers of medical education and CPD programs have adopted e-learning approaches, which have been shown to provide flexible, low-cost, user-centered, and easily updated learning. The effectiveness of e-learning varies from context to context, however, and has also been shown to make considerable demands on users' motivation and "digital literacy" and on providing institutions. Consequently, there is a need to evaluate the effectiveness of e-learning in healthcare as part of ongoing quality improvement efforts. This article outlines the key issues for developing successful models for analyzing e-health learning.

Avoiding pathologies in very deep networks

Choosing appropriate architectures and regularization strategies of deep networks is crucial to good predictive performance. To shed light on this problem, we analyze the analogous problem of constructing useful priors on compositions of functions. Specifically, we study the deep Gaussian process, a type of infinitely-wide, deep neural network. We show that in standard architectures, the representational capacity of the network tends to capture fewer degrees of freedom as the number of layers increases, retaining only a single degree of freedom in the limit. We propose an alternate network architecture which does not suffer from this pathology. We also examine deep covariance functions, obtained by composing infinitely many feature transforms. Lastly, we characterize the class of models obtained by performing dropout on Gaussian processes.

An extension of slow feature analysis for nonlinear blind source separation

We present and test an extension of slow feature analysis as a novel approach to nonlinear blind source separation. The algorithm relies on temporal correlations and iteratively reconstructs a set of statistically independent sources from arbitrary nonlinear instantaneous mixtures. Simulations show that it is able to invert a complicated nonlinear mixture of two audio signals with a high reliability. The algorithm is based on a mathematical analysis of slow feature analysis for the case of input data that are generated from statistically independent sources. © 2014 Henning Sprekeler, Tiziano Zito and Laurenz Wiskott.

An energy-efficient hopping robot based on free vibration of a curved beam

This paper explores a design strategy of hopping robots, which makes use of free vibration of an elastic curved beam. In this strategy, the leg structure consists of a specifically shaped elastic curved beam and a small rotating mass that induces free vibration of the entire robot body. Although we expect to improve energy efficiency of locomotion by exploiting the mechanical dynamics, it is not trivial to take advantage of the coupled dynamics between actuation and mechanical structures for the purpose of locomotion. From this perspective, this paper explains the basic design principles through modeling, simulation, and experiments of a minimalistic hopping robot platform. More specifically, we show how to design elastic curved beams for stable hopping locomotion and the control method by using unconventional actuation. In addition, we also analyze the proposed design strategy in terms of energy efficiency and discuss how it can be applied to the other forms of legged robot locomotion. © 1996-2012 IEEE.

Physical connection and disconnection control based on hot melt adhesives

Physical connection and disconnection control has practical meanings for robot applications. Compared to conventional connection mechanisms, bonding involving a thermal process could provide high connection strength, high repeatability, and power-free connection maintenance, etc. In terms of disconnection, an established bond can be easily weakened with a temperature rise of the material used to form the bond. Hot melt adhesives (HMAs) are such material that can form adhesive bonds with any solid surfaces through a thermally induced solidification process. This paper proposes a novel control method for automatic connection and disconnection based on HMAs. More specifically, mathematical models are first established to describe the flowing behavior of HMAs at higher temperatures, as well as the temperature-dependent strength of an established HMA bond. These models are then validated with a specific type of HMA in a minimalistic robot setup equipped with two mechatronic devices for automated material handling. The validated models are eventually used for determining open parameters in a feedback controller for the robot to perform a pick-and-place task. Through a series of trials with different wooden and aluminum parts, we evaluate the performance of the automatic connection and disconnection methods in terms of speed, energy consumption, and robustness. © 1996-2012 IEEE.

Linear multimodal actuation through discrete coupling

Due to technological limitations, robot actuators are often designed for specific tasks with narrow performance goals, whereas a wide range of behaviors is necessary for autonomous robots in uncertain complex environments. In an effort to increase the versatility of actuators, we introduce a new concept of multimodal actuation (MMA) that employs dynamic coupling in the form of clutches and brakes to change its mode of operation. The dynamic coupling allows motors and passive elements such as springs to be engaged and disengaged within a single actuator. We apply the concept to a linear series elastic actuator which uses friction brakes controlled online for the dynamic coupling. With this prototype, we are able to demonstrate several modes of operation including stiff position control, series elastic actuation as well as the possibility to store and release energy in a controlled manner for explosive tasks such as jumping. In this paper, we model the proposed concept of actuation and show a systematic performance analysis of the physical prototype that we developed in our laboratory. © 1996-2012 IEEE.

Linear dimensionality reduction: Survey, insights, and generalizations

© 2015 John P. Cunningham and Zoubin Ghahramani. Linear dimensionality reduction methods are a cornerstone of analyzing high dimensional data, due to their simple geometric interpretations and typically attractive computational properties. These methods capture many data features of interest, such as covariance, dynamical structure, correlation between data sets, input-output relationships, and margin between data classes. Methods have been developed with a variety of names and motivations in many fields, and perhaps as a result the connections between all these methods have not been highlighted. Here we survey methods from this disparate literature as optimization programs over matrix manifolds. We discuss principal component analysis, factor analysis, linear multidimensional scaling, Fisher's linear discriminant analysis, canonical correlations analysis, maximum autocorrelation factors, slow feature analysis, sufficient dimensionality reduction, undercomplete independent component analysis, linear regression, distance metric learning, and more. This optimization framework gives insight to some rarely discussed shortcomings of well-known methods, such as the suboptimality of certain eigenvector solutions. Modern techniques for optimization over matrix manifolds enable a generic linear dimensionality reduction solver, which accepts as input data and an objective to be optimized, and returns, as output, an optimal low-dimensional projection of the data. This simple optimization framework further allows straightforward generalizations and novel variants of classical methods, which we demonstrate here by creating an orthogonal-projection canonical correlations analysis. More broadly, this survey and generic solver suggest that linear dimensionality reduction can move toward becoming a blackbox, objective-agnostic numerical technology.

Manopt, a matlab toolbox for optimization on manifolds

Optimization on manifolds is a rapidly developing branch of nonlinear optimization. Its focus is on problems where the smooth geometry of the search space can be leveraged to design effcient numerical algorithms. In particular, optimization on manifolds is well-suited to deal with rank and orthogonality constraints. Such structured constraints appear pervasively in machine learning applications, including low-rank matrix completion, sensor network localization, camera network registration, independent component analysis, metric learning, dimensionality reduction and so on. The Manopt toolbox, available at www.manopt.org, is a user-friendly, documented piece of software dedicated to simplify experimenting with state of the art Riemannian optimization algorithms. By dealing internally with most of the differential geometry, the package aims particularly at lowering the entrance barrier. © 2014 Nicolas Boumal.