21793 research outputs found
Sort by
Collaborative repair as dealienation: an exploration of degrowth technology practice
Get PDFThe degrowth hypothesis could be summarised by the following: even if limitless growth were biophysically possible—which it is almost certainly not—it would not be desirable. The degrowth project bills itself as more than just critique: it is “a normative concept with analytical and practical applications” (Kallis et al 2018). Yet while scholarship has meaningfully engaged with ecological economics and political ecology to interrogate the metabolic imbalances and distributional asymmetries of growth-centric society, empirical investigation of alternative “living degrowth” are rare (Brossman and Islar 2020). Degrowth research focusing on questions of technological normativity, or “technology practice” (Drengson 1995) are few and mostly limited to work adopting largely quantitative, or metabolic, approaches to technology, for instance in the ‘low-tech’ movement. These predominantly biophysical framings are clearly necessary in apprehending, and acting upon, the impossibility of endless growth and commodity innovation/production. They are however, less adequate in accounting for the undesirability of endless growth and material accelerations, and in indicating new, more desirable pathways for technology practice moving forward.
The present empirical study consists in first-person observation and interviews carried out in a Montreal amateur repair community in 2021-2022. The phenomenon of collaborative repair, or Repair Cafés, is a practice geared to the downscaling of material throughput through the collectivisation of tools, space and repair knowledge. Through observation and analysis, a cluster of questions was asked: how could we begin to think about degrowth technology practice? What would it look like? Can the features of collaborative repair offer us hints? Drawing on recent scholarly efforts to revive ‘alienation’ as a valid theme for social inquiry, and in addressing the noted need for degrowth to think more seriously about “dealienation” (Brownhill et al 2012), the present study looks at collaborative repair as a testing site for the suitability of these concepts, and for their potential application in a proposed degrowth research mandate focused on technology practice. This study is founded on a methodological conviction that when one engages in practice, one not only does something, one also understands that one is doing something, inevitably investing the action with meaning (Jaeggi 2018). From this point of view, and beyond metabolic and redistributive ends, collaborative repair effects a rehabilitation of meaningful subject-subject, subject-time and subject-object relations—relations typically characterised by alienation in industrial commodity economies. The present study also recommends that degrowth think seriously about “resonance” (Rosa 2019) as a more useful and coherent alternative to ‘autonomy’ when conceptualising alienation’s ‘other’. Such a framing appears critical for both elaborating a degrowth critique of technology and enriching discussions of how degrowth normativity bears on practice.
Application of Attention Mechanism in Deep Neural Network Architecture for System Failure Prognostics
Get PDFMachine health monitoring and management are essential improvements that must be considered in the industry toward smart manufacturing. Intelligent prognosis and health management (PHM) systems have demonstrated remarkable capabilities for industrial use and, consequently, have become active research areas in the last several decades. Predictive Maintenance (PM) generally predicts faults or breakdowns in a deteriorating system to optimize maintenance efforts by evaluating the system's status using historical data. In this strategy, the Remaining Useful Life (RUL) of the components is anticipated using characteristics, which typically include sensors and operational profiles. This research aims to evaluate the possibility of predicting the RUL of a system based on sensor data by deploying an attention-based deep learning model. RUL prediction based on the attention mechanism is a relatively new approach with promising results. One advantage of this approach is that it can be useful for interpreting the results and understanding the underlying factors contributing to the RUL. Applying an attention mechanism to find temporal dependencies also shows improvement in model performance by detecting the most important part of the sequences to be passed to the prediction model. Our proposed model has shown a noticeable impact on the performance of the neural network architecture from the attention mechanism added to the pipeline by keeping the model light in terms of computational resources and training time. The proposed model clearly shows the attention mechanism's high impact on predicting sequential data. This technique can also be used in more complex ensemble-based architectures to improve performance
Frost Damage of Internally Insulated Retrofitted Solid Brick Walls: Experimental Work and Hygrothermal Modeling
Get PDFFrost damage is one of the most common deteriorations in porous clay bricks. Furthermore, adding internal thermal insulation (ITI) to a building envelope increases its thermal resistance and decreases conductive heat loss; however, it may cause durability performance issues of the wall in cold areas. Although numerous previous studies have examined the addition of ITI, there is still a need for additional research on this subject that considers other factors, including brick properties (BP), wind-driven rain (WDR) exposure levels, and insulation thicknesses and types. Considering exposure level, this study investigates frost durability (FD) and BP of five brick types and their effect on the retrofitting process. This study is divided into two sections: the first is to predict the FD of five brick types using three different methods: Canadian standard CSA A82.1, critical degree of saturation (Scrit), and durability factor (Df); the second is coupling Scrit measurements and BP with predictive modelling hygrothermal performance to assess the risk of freeze damage using hygrothermal simulation program WUFI Pro 6.5. The results revealed that the five brick types presented interesting variations between their properties, probably due to the effects of the service life for old brick types, where it has been manufactured and built for hundreds of years, the manufacturing process for new bricks, and heterogeneity between samples in terms of pore size distribution. Therefore, comprehending the influence of BP on the frost resistance (FR) of clay bricks plays a fundamental role in controlling the FD phenomena and avoiding the deterioration of the clay bricks, particularly after adding ITI. A good correlation was found between 5-h BWA, C/B, A-value, and 24-h CWA. The Scrit is around 55%, 50%, 50, 65%, and 35% of ERP, ERU, ENB, ENO, and IRM, respectively, ±5%. An empirical formula was developed to determine Scrit based on 24-h CWA, 5-h BWA, and compressive strength. The modelling showed that depending on the BP and moisture exposure level, adding ITI increases frost damage (FD) mainly in the second layer (i.e., 15 mm inwards from the outer side) and the middle layer and adding ITI did not impact walls made with samples having a Scrit of 0.60 or higher, even when subjected to higher WDR. In contrast, clay bricks with a Scrit of 0.55 or less require additional attention, particularly when exposed to higher WDR. Furthermore, depending on the 24-h CWA standard limit may provide a good result for preventing FD after adding ITI; however, the 24-h CWA of a brick is not an absolute indicator to predict FR. In addition, the insulation types and level do not impact FD for the wall made with brick samples with a Scrit of 50% and higher. Although correlations in the BP were found in the experimental and the modelling phases, a certain percentage of bricks are likely to fail, given the wide variation in properties
Retailing in College Towns: Spatial Location and Multimodal Commuting
No full textCollege town retailers face specific challenges given their unique context. Our study contributes to research on retailing by exploring the college town context to generate new insights while also explicating the impact of store-university distance and commuting multimodality. Our research questions are: (1) how does store-university distance moderate the relationship between university foot traffic and retailer foot traffic? (2) How does university foot traffic interact with multimodal commuting in affecting retailer foot traffic? (3) What moderating effects did changes in the containment and health responses during the COVID-19 pandemic exert upon these relationships? Using foot traffic data from 157 Walmart and Target stores in 38 college towns from 2018-2020, our study indicates that university foot traffic has a positive impact on store visits. Moreover, the positive effect is weakened as the store to university distance increases but is strengthened with greater commuting multimodality. In addition, we find that the intensity of pandemic related containment and health measures amplifies both the negative effect exerted by the distance between the store and university and the positive effect exerted by commuting multimodality upon the positive correlation between university visits and store visits. This research provides information that can be utilized by government officials and retail managers to enhance consumer accessibility to retail stores in college towns and to inform strategies to respond and recover from a pandemic such as COVID-19
On the Development of Praseodymium-Doped Radioluminescent Nanoparticles and Their Use In X-ray Mediated Photodynamic Therapy Of Glioblastoma Cells
Get PDFDespite decades of research, few advancements have been made toward improving the prognosis of patients with glioblastoma, a lethal and invasive form of brain cancer. The current standard of care is fluorescence-guided surgical resection followed by radiotherapy and chemotherapy. Fluorescence guided surgery is performed using 5-aminolevulinic acid (5-ALA), a prodrug that induces the accumulation of fluorescent protoporphyrin IX (PPIX) in malignant cells. Conveniently, 5-ALA mediated production of PPIX is also renowned as the most popular photodynamic therapy (PDT) agent in the world. PDT is a treatment that uses visible light to stimulate a photosensitizer to produce reactive oxygen species, which can damage and kill cells. However, the technique is limited by the tissue depth penetration of light. The advent of nanomedicine has enabled the possibility to achieve PDT by using luminescent nanoparticles to alter the incident excitation source. When X-rays are used to excite the nanoparticles, the process is called X-ray mediated photodynamic therapy (X-PDT).
Herein, we have developed NaLuF4:Pr3+ radioluminescent nanoparticles to achieve X-PDT. The emission spectrum of Pr3+ exhibits strong spectral overlap with the absorption spectrum of PPIX, an endogenous photosensitizer. A reproducible route to synthesizing uniform NaLuF4:Pr3+ nanoparticles at sizes relevant for cell uptake was developed, and the spectroscopic properties of the nanoparticles were evaluated prior to in vitro studies. The nanoparticles were found to exhibit persistent luminescence, and a mechanism was developed to explain the charge (de-)trapping process. The nanoparticle composition was optimized for excitation of PPIX and studied in the human glioblastoma cell line called U251. We evaluated the therapeutic effect of the nanoparticles with and without 5-ALA to establish the radiosensitization capability as well as the X-PDT effect. Three nanoparticle concentrations were studied using 4 radiation doses, including those relevant for intraoperative radiotherapy which is performed on the tumor cavity immediately after resection. The effects on stress, death, damage, senescence and proliferation were studied. and demonstrate promising results at a proof-of-concept level. Throughout this work, current clinical practice guided our experimental design, providing a strong foundation toward using Pr3+-doped nanoparticles for X-PDT in an intraoperative setting using endogenous PPIX as the photosensitizer
An Integrated Data-Driven Failure Prediction and Risk Management Approach for Water Mains
Get PDFWater distribution networks (WDNs) play a vital role in reliably delivering clean potable water to the society. The deterioration of water infrastructures that have drastically increased throughout the major urban centers has caused increasing water-main (WM) failures with severe consequences such as disruption of services and revenue losses. Effective management of WMs is essential. This involves repairing the WMs and implementing strategies to minimize water loss. Models should be used to predict breaks ahead of their occurrences and to plan rehabilitation. The use of these models would promote sustainable infrastructures and save costs. This research integrates the probability of failure (POF) derived from a random forest failure prediction model (predictive analytics) with a risk management strategy (prescriptive analytics) in a cold region in Canada. To investigate the effect of environmental factors, freezing index was considered and found to be among the top three most important attributes. In the proposed predictive analytics process Principal Component Analysis (PCA) was implemented for data reduction. Clustering is applied to avoid under/overestimating WM failure prediction and find the most similar cohorts. The results outlined that clustering considerably improved the prediction and risk-analysis outcomes. Finally, with the proposed risk management strategy, results showed that 3.68% of the network's total length is at high risk, and needs immediate action for fixing; however, it is only 0.07 to 1.02% of the network's total length when clustering was performed. Therefore, there was a 67–80% improvement in having WMs with high-rating risk compared to when no clustering was performed
Structural Response Assessment of Timber Log Walls Under In-Plane Lateral Load
No full textThis research is a comprehensive investigation into the factors contributing to the lateral resistance force and initial stiffness of log walls in log houses. Contributing factors include the type of log used, the method of bonding the corners, the presence of geometric irregularities, the effect of vertical loads, and the coefficient of friction between logs. The study uses a finite element modelling approach to evaluate the effect of these factors on the lateral resistance force and initial stiffness of log walls. Once the model is developed, small- and wall-scale experiments validate the model. Parametric studies are conducted to evaluate the effects of the reinforcement systems with different steel rod configurations on the log wall resistance force and initial stiffness. Various parametric studies are presented assessing the effect of the penetration length, openings, friction, pre-compression load, aspect ratio, reinforcement methods, log profiles and wall’s orthogonal joineries on the lateral strength and stiffness of the wall systems.
The findings of this study provide insights into ways to improve the design and construction of log houses, enhancing their lateral resistance capabilities. This, together with increasing the safety and stability of log houses, facilitates the building industry with a deeper understanding of the lateral behaviour of log walls with different bonded corners
Investigating and Testing Performance Issues in Deep Learning Frameworks
Get PDFMachine Learning (ML) and Deep Learning (DL) applications are becoming more popular due to the availability of DL frameworks such as PyTorch, Keras, and TensorFlow. Therefore, the quality of DL frameworks is essential to ensure DL/ML application quality. Given the computationally expensive nature of DL tasks (e.g., training), performance is a critical aspect of DL frameworks. However, optimizing DL frameworks may have its own unique challenges due to the peculiarities of DL (e.g., hardware integration and the nature of the computation).
In this thesis, we first aim to better understand performance bugs in DL frameworks by conducting an empirical study. We conduct our study on PyTorch and TensorFlow by mining and studying their performance and non-performance bug reports from their respective GitHub repositories. We find that 1) the proportion of newly reported performance bugs increases faster than fixed performance bugs, and the ratio of performance bugs among all bugs increases over time; 2) performance bugs take more time to fix, have larger fix sizes, and more community engagement (e.g., discussion) compared to non-performance bugs; and 3) we manually derived a taxonomy of 12 categories and 19 sub-categories of the root causes of performance bugs in DL frameworks by studying all performance bug fixes.
We then aim to investigate the potential of differential testing as a viable technique to detect and prevent performance bugs in DL frameworks. To do so, we train and evaluate two state-of-the-art CNN and RNN architectures (i.e., the Lenet-5 architecture on the MNIST dataset and the LSTM architecture on the IMDB movie review dataset), using different DL frameworks (i.e., PyTorch, Keras, and TensorFlow), and different configurations (i.e., the training dataset sample size, the batch size, the number of epochs, the weight initialization technique, the data type, the hardware used, the learning rate, and the dropout rate). To assess the performance of the DL models, we use a variety of performance metrics (i.e., training/inference time, hardware (CPU or GPU) usage during training/inference, and memory (RAM or GPU VRAM) usage during training/inference). Then, we compare the performance of the DL models across the DL frameworks. We train and evaluate 21,870 Lenet5 models and 21,870 LSTM models across the DL frameworks, for a grand total of 43,740 models. Our experiments took over 42 days. We find that 1) differences in performance between different DL frameworks, for the same task, may be indicative of a performance optimization opportunity/performance bug; 2) our approach is viable when training and evaluating a smaller number of DL models, which makes it more accessible for developers.
Finally, we present some potential avenues for future work that aim to further study performance bugs in DL frameworks
SMEs and Sustained Growth During Prolonged Crisis
Get PDFThe main purpose of this dissertation is to investigate the process by which small and medium-sized enterprises achieve sustained growth in crisis environments. It consists of three essays that explore this process from theoretical, empirical and normative perspectives, and elaborates the choices and actions taken by managers toward sustained growth in a crisis environment. The focus is on the process of internationalization and the development of versatile resources and capabilities that contribute to SME sustained growth. Previous studies have not adequately explored sustained growth over a prolonged crisis period nor the key driving forces required to achieve it. This study makes a unique contribution by conducting a process analysis and providing a holistic method to capture the internationalization and sustained growth of SMEs in a crisis context. The theoretical foundations of Penrose’s growth theory and the dynamic capabilities literature help confirm the relationship between proactive internationalization, capability development, and growth, and shed light on how this relationship is achieved and endures in a crisis context. Theoretical contributions lie in exposing our assumptions in the crisis literature and shifting the mode of theorizing to a process-based approach, demonstrating that sustained growth can be achieved in a crisis context. The study identifies a multidimensional framework that uncovers a continuous, non-linear process of versatile resource and capability development that contributes to sustained growth, through specific managerial and value creating mechanisms
Date Palms as Living Infrastructure: An Ethnography of Human-Plant Relationship in Bam City, Iran
Get PDFThis ethnography narrates the story of date palms as a living infrastructure in supporting Bam gardeners as they navigate the daily uncertainties in Iran. The research focuses on two main areas. The first part examines the intimate gardener-date palm relationship, exploring how this intimate bond is nurtured through their daily interactions and the profound embodiment of the tree. This multi-species bonding serves as a source of inspiration, instilling a sense of hope and resiliency within the gardening community, especially during challenging times. To illustrate this point, the research offers a poignant example of the 2003 destructive earthquake, revealing how the date palm gardening infrastructure played a pivotal role in the recovery of Bam gardeners.
In the second part, the research contextualizes this bonding within Iran’s broader socio-political landscape. It sheds light on how the resilient date palm infrastructure emerged as a result of the land and water reform policies following the 1979 Islamic revolution