Over the last decades we have witnessed a profound digitalization of tangible products. While this shift offers great opportunities, it also exposes product developing industries to significant challenges. In these industries organizations, markets, and technologies are tuned for mass production, providing competitive advantage through scale economics. Typically, firms exercise modular strategies to deliver such scale benefits. Rooted in Herbert Simon’s notion of near decomposability, modular product architectures allow for production assets, such as tools, processes, and plants, to be effectively reused across product variants and over generations of designs. However, they come at a price; modularity requires overall design specifications to be frozen well before production. In practice, this tends to inscribe functional purpose in the structures of the system, effectively preventing firms from taking advantage of the speed by which digitized products can be developed and modified. The main objective of this thesis is to investigate and explain how product developing organizations adapt architectural thinking to balance the proven benefits of modularity and the emerging opportunities provided by digital technology. In doing so, it introduces a complementary architectural frame, grounded in Christopher Alexander’s seminal work on patterns. This frame associates the concept of architecture with generativity and reuse of ideas, rather than scale economics and reuse of physical assets. Sensitizing the theoretical framework through a longitudinal case study of digital product innovation this thesis derives several implications for theory and practice. Across four embedded cases in the automotive industry it demonstrates that generative capability follows from a shared organizational view on products as enablers and catalyzers of new, yet unknown functionality. Such an emergence-centric view requires product developing firms to rethink existing governance models. Rather than exercising control through specific functionality, inscribed in modular product structures, it offers the benefit of influencing innovation through general functional patterns, serving as raw material in distributed and largely uncoordinated innovation processes. This shift in focus, from specific functionality to general functional patterns, enables a new strategic asset for product developing firms. It opens up for proactive rather than reactive strategies, where the architecture makes an instrument to cultivate new ideas and business opportunities, rather than a tool for cost savings.

Thesis (Ph. D.)--University of Missouri-Columbia, 2007. / The entire dissertation/thesis text is included in the research.pdf file; the official abstract appears in the short.pdf file (which also appears in the research.pdf); a non-technical general description, or public abstract, appears in the public.pdf file. Title from title screen of research.pdf file (viewed on February 13, 2008) Vita. Includes bibliographical references.

Thesis (Ph.D.)--University of Wollongong, 2004. / Typescript. Bibliographical references: leaf [359]-377.

Thesis (M.S.)--Missouri University of Science and Technology, 2009. / Vita. The entire thesis text is included in file. Title from title screen of thesis/dissertation PDF file (viewed April 27, 2009) Includes bibliographical references.

Thesis (M.S.)--Brigham Young University. Dept. of Mechanical Engineering, 2007. / Includes bibliographical references (p. 101-103).

Thesis (Master)--İzmir Institute of Technology, İzmir, 2006 / Keywords: Modular design, mass production, mas customization, customization. Includes bibliographical references (leaves 96-99).

The history of total knee arthroplasty stretches back over 70 years. Many studies have shown that TKA is, in general, a successful operation for the relief of joint pain, with patient satisfaction rates of 90-95% and implant survival rates at 10-15 years of greater than 90%. However, a number of studies have also shown the potential for failures or complications arising post-implantation leading to revision surgery. This thesis presents finite element (FE) models of the distal femur following primary and revision total knee arthroplasty. Pre-arthroplasty models are also developed for comparison. Particular attention is given to how femoral component design and method of fixation impacts the mechanical environment of the distal femur and stability of the prosthesis. FE analyses with fully bonded interfaces indicate that femoral components are subject to areas of low stress (stress shielding) immediately under the anterior flange and chamfer regardless of internal implant features. However, internal implant features were found to play a role in the pattern and magnitude of stress concentrations. Both stresses and motions were observed to increase with increasing flexion angle, indicating the importance of testing at multiple angles. The initial models of the distal femur were extended to incorporate the effects of ageing and endosteal thinning of the femoral cortex, through novel application of pre-existing FE modelling techniques, specifically the ability to assign variable material properties corresponding to the nodal temperatures output from a heat transfer analysis. The findings from this study indicate that older patients with osteoporosis may be at increased risk of periprosthetic fracture compared to younger healthy patients. The use of a revision femoral component with a cemented stem as a means to mitigate this fracture risk was also investigated. FE analyses using frictional interfaces were employed to determine the influence of femoral component design on micromotion at the interface. These models showed that all primary implants were subject to similar magnitudes of relative motion at the interface, however, the distinct internal implant features led to very different regional variations. Furthermore, certain internal implant features (i.e. femoral box) were found to be highly sensitive to errors in surgical bone cuts. This aspect of the thesis also concluded that the addition of a stem served to significantly reduce motions at the interface in comparison to primary stemless implants. Long stemmed prostheses were found to result in the smallest levels of interface motion. This study also detailed the design and creation of an in vitro test setup for the purposes of determining the influence of stem length and fixation on the stability of revision prostheses. Experimental results using this test rig showed that a cemented short stem provides as much initial stability as the uncemented long stem, and is easier to fit surgically. Corresponding FE models incorporating a virtual representation of the test rig and in vitro loading conditions revealed that the relative motion at the multi-planar bone-prosthesis interface cannot be adequately described using a single reference point. However, in vitro setups may be used to predict a general measure of implant stability and to provide a source of calibration for FE. The distal femur models were further modified to investigate how the presence of condylar defects as classified by AORI defect classification system (Engh 2006) and weak osseous support due to osteoporosis may adversely affect the survival of the prosthesis. These investigations revealed that fixation of the femoral component, the presence of a large condylar defect and the level of osseous support all had an impact on stress in the implant, it is concluded that a non-modular approach should be adopted in older patient groups with severe osteoporosis to mitigate the risk of component junction failure and distal femoral fracture.

A subfamília de roedores Sigmodontinae representa o clado com a maior diversidade e distribuição de mamíferos na região neotropical, sendo que a maior parte das espécies são endêmicas da América do Sul. Com inúmeras diferenças ecomorfológicas, o padrão de diversificação desses roedores, por um ponto de vista biogeográfico e filogenético, tem sido bastante debatido. Nesta tese, busco compreender a evolução dos caracteres cranianos dos sigmodontíneos, a partir do arcabouço teórico da genética quantitativa e integração morfológica. Dessa forma, tive como objetivo geral avaliar os padrões e as magnitudes de integração morfológica para compreender a associação dos caracteres e explorar suas potenciais consequências evolutivas no crânio dos Sigmodontinae. A partir de um banco de dados contendo 2897 indivíduos de 39 espécies da subfamília, testei a similaridade estrutural das matrizes de correlação e covariância ao compará-las entre todos os táxons medidos (representados por 35 medidas cranianas). Avaliei também se a história evolutiva do grupo teve influência sobre os padrões da estrutura de covariância fenotípica. Além disso, testei a presença de módulos no crânio desses roedores, a partir das hipóteses de desenvolvimento e função comum nos crânios dos mamíferos. Por fim, simulei seleção natural nesses crânios a fim de investigar as possibilidades evolutivas na associação dos caracteres cranianos nos diferentes táxons da subfamília. De uma maneira geral, os sigmodontíneos apresentaram um padrão das relações dos caracteres muito semelhante, enquanto a magnitude em que esses caracteres estão integrados variou bastante entre as espécies sendo que a amplitude dessa variação foi próxima a que já foi detectada entre outras ordens de mamíferos eutérios. Além disso, tanto os padrões quanto as magnitudes obtidas não estiveram, até um certo ponto, associadas a história evolutiva (filogenia) desses roedores. Dessa forma, observa-se que após um período de diversificação de aproximadamente 12 milhões de anos, a plasticidade na magnitude de integração acoplada ao fato que esta magnitude nunca se aproxima de 1 (portanto com algum grau de flexibilidade) pode fornecer uma explicação sobre como a grande diversidade morfológica craniana surgiu neste grupo de mamíferos, mesmo com a grande conservação nos padrões de integração. Mais ainda, os sigmodontíneos compartilham um padrão de modularidade craniana comum entre a maioria das espécies, relacionada com as hipóteses funcionais e de desenvolvimento testadas. O padrão de modularidade mostrou-se influenciado pela variação de tamanho, associado ao primeiro componente principal (CP1) de todas as espécies, assim como a magnitude geral de integração do crânio. Quanto maior a variância alométrica no CP1 maior a magnitude de associação entre os caracteres do crânio e, dessa maneira, menos modular é o crânio. Essa relação é importante para compreender o potencial da resposta evolutiva, pois, independente da direção da pressão seletiva, espécies com maior magnitude geral de integração dos caracteres são mais restritas evolutivamente, ou seja, tendem a responder à seleção na direção do eixo de maior variação (tamanho). Em contrapartida, espécies com menores magnitudes entre os caracteres cranianos são mais flexíveis a responder na direção em que seleção está atuando. Interessantemente, os sigmodontíneos apresentaram potenciais evolutivos tanto próximo dos mamíferos mais flexíveis (primatas e morcegos) quanto daqueles que possuem os maiores índices de restrição (marsupiais) / The Sigmodontinae subfamily of rodents represents the clade with the greatest diversity and distribution among mammals in the Neotropical region, with most of the species endemic to South America. With numerous ecomorphological differences, the diversification pattern of these rodents, from a phylogenetic and biogeographic point of view, has been intensely debated. In this thesis, I seek to understand the evolution of the cranial traits of the sigmodontine, using the theoretical framework of quantitative genetics and morphological integration. Thus, I aimed at evaluating the patterns and magnitudes of morphological integration, to understand the association of traits and to explore the potential evolutionary consequences of these associations for the Sigmodontinae skull. Using a collected database containing 2897 individuals of 39 species of the subfamily, I tested the structural similarity of covariance and correlation matrices by comparing them between all measured taxons (represented by 35 cranial measurements). I also evaluated if the evolutionary history of the group had an influence on the patterns of phenotypic covariance structure. Furthermore, I tested the presence of modules in these rodents skulls, employing shared developmental and functional hypothesis proposed for mammalian skulls. Finally, I simulated natural selection in these skulls in order to investigate the evolutionary possibilities in the association of cranial traits in different taxa of the subfamily. In general, the sigmodontine had very similar patterns of traits relationship, while the magnitude of trait association varied greatly among species, and the amplitude of this variation was close to what has already been detected within other Eutheria mammalian orders. In addition, both the patterns as well as the magnitudes obtained were not, to a large extent, associated with the evolutionary history (phylogeny) of these rodents. Thus, I observed that after a period of diversification of approximately 12 million years, this magnitude of integration plasticity coupled with the fact that the magnitude of association is never 1 (granting some degree of flexibility), may provide an explanation for how the wide diversity in cranial morphology appeared in this mammalian group, even with the wide conservation in the integration patterns. Moreover, the sigmodontine share a common pattern of cranial modularity among most species, which are related to the tested functional and developmental hypotheses. The modularity pattern appears to be influenced by variation in size, associated with the first principal component (PC1) of all species, as well as the overall magnitude of skull integration. The greater the variance in the allometric PC1, the greater the association between the skull traits and, thus, the skull is less modular. This relationship is important to understand the potential of the evolutionary response, seeing that, independent of the direction of selective pressure, species with higher general integration magnitude are more evolutionarily constrained, i.e, tend to respond to selection in the direction of the axis with the greatest variation (size). In contrast, species with smaller magnitudes between cranial characters are more flexible to respond in the direction in which selection is acting. Interestingly, the sigmodontine exhibited a potential to evolutionary responses that range from much flexible, both close to mammals (primates and bats ), as well as with those who have the highest levels of restriction (marsupials)

Made available in DSpace on 2014-06-12T15:57:56Z (GMT). No. of bitstreams: 2 arquivo3237_1.pdf: 1530844 bytes, checksum: dcdb6221a7c974cbfc9e96c7629001ef (MD5) license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2010 / Conselho Nacional de Desenvolvimento Científico e Tecnológico / Atualmente, arquitetura baseada em componentes é a abordagem mais utilizada no desenvolvimento de softwares complexos; esta tem como principal objetivo a atribuição dos requisitos da aplicação aos componentes. Uma das técnicas mais difundidas para especificação de requisitos é a utilização de Casos de Uso. Em geral, arquiteturas de software baseadas em componentes resultam em implementações onde o código relativo a um caso de uso está espalhado e entrelaçado em diversos componentes do sistema, caracterizando um crosscutting concern. Isto ocorre porque técnicas tradicionais, como Orientação a Objetos (OO), não oferecem mecanismos que sejam capazes de modularizar este tipo de concern. Recentemente, novas técnicas de modularização como aspectos, mixins e classes virtuais, foram propostas para tentar resolver este problema. Estas técnicas podem ser usadas para agrupar o código relacionado a um único caso de uso em uma nova unidade de modularização. Este trabalho analisa qualitativa e quantitativamente o impacto causado por este tipo de modularização de casos de uso. Nós exploramos duas técnicas baseadas em Orientação a Aspectos (OA): (i) Casos de Uso como Aspectos - onde utilizamos os construtores de AspectJ para isolar todo código relativo à implementação de um caso de uso em um aspecto; e (ii) Casos de Uso como Colaborações Plugáveis - onde usamos os construtores de CaesarJ para modularizar implementações de casos de uso através de uma composição hierárquica de colaborações. Nós executamos dois estudos de casos onde comparamos as implementações OA de casos de uso com sua implementação OO. No processo de avaliação extraímos métricas tradicionais e contemporâneas incluindo coesão, acoplamento e separação de concerns e analisamos modularidade em termos de atributos de qualidade de software como: plugabilidade, rastreabilidade e suporte para desenvolvimento em paralelo. Nossos resultados indicam que modularidade é um conceito relativo e sua análise depende de outros fatores além do sistema alvo, das métricas e da técnica aplicada

Many of the algorithms used for community detection in temporal networks have been adapted from static network theory. A common approach in dealing with the temporal dimension is to create multiple static networks from one temporal, based on a time condition. In this thesis, focus lies on identifying the optimal partitioning of a few temporal networks. This is done by utilizing the popular community detection algorithm called Generalized Louvain. Output of the Generalized Louvain comes in two parts. First, the created community structure, i.e. how the network is connected. Secondly, a measure called modularity, which is a scalar value representing the quality of the identified community structure. The methodology used is aimed at creating a comparable result by normalizing modularity. The normalization process can be explained in two major steps: 1) study the effects on modularity when partitioning a temporal network in an increasing number of slices. 2) study the effects on modularity when varying the number of connections (edges) in each time slice. The results show that the created methodology yields comparable results on two out of the four here tested temporal networks, implying that it might be more suited for some networks than others. This can serve as an indication that there does not exist a general model for community detection in temporal networks. Instead, the type of network is key to choosing the method.