48 research outputs found
Improving operational efficiency of a semiconductor equipment manufacturing warehouse through strategic allocation of parts
No full textThesis: M. Eng. in Manufacturing, Massachusetts Institute of Technology, Department of Mechanical Engineering, 2015.Cataloged from PDF version of thesis.Includes bibliographical references (page 75).The work addresses the operational inefficiency problem in a semiconductor equipment manufacturing warehouse of Applied Material's Varian Semiconductor Business Unit. At Varian, the target part delivery time from the warehouse to the production floor is 24 hours. However, during busy periods, parts are not delivered on time. Late part delivery from the warehouse to the production floor could delay the machine laydown date, which in turn could result in late or missed shipment of tools to the customers, which can be very costly. To improve the efficiency and the reliability of the warehouse, picking efficiency is to be improved. Parts from the warehouse are picked from three picking locations- Vertical Lift Modules (VLMs), GL, and RK. VLMs are automated machines, while GL and RK are manual picking zones. Picking an order from GL takes the most amount of time. The overall picking efficiency at the warehouse can be improved by partially shifting the workload from GL to the VLMs, and by further improving the picking efficiency at the VLMs. The workload from GL to the VLMs is shifted by transferring fast moving parts from GL to the VLMs. The picking efficiency of the VLMs is improved by balancing the workload of all five VLM pods, and by employing a more efficient 'pick-and-consolidate' picking strategy. The workload at GL is decreased by 25% and the workload at VLMs is increased by 13%. Despite the increase in workload at VLMs, 23% time savings could be achieved by balancing the utilization of all five VLM pods. Additional time savings of 20 minutes per order (8%) could be achieved by using 'pick-and-consolidate' picking strategy over 'pick-and-pass' picking strategy.by Paramveer Singh Toor.M. Eng. in Manufacturin
Advances in spectral learning with applications to text analysis and brain imaging
Get PDFSpectral learning algorithms are becoming increasingly popular in data-rich domains, driven in part by recent advances in large scale randomized SVD, and in spectral estimation of Hidden Markov Models. Extensions of these methods lead to statistical estimation algorithms which are not only fast, scalable, and useful on real data sets, but are also provably correct. Following this line of research, we make two contributions. First, we propose a set of spectral algorithms for text analysis and natural language processing. In particular, we propose fast and scalable spectral algorithms for learning word embeddings -- low dimensional real vectors (called Eigenwords) that capture the “meaning” of words from their context. Second, we show how similar spectral methods can be applied to analyzing brain images. State-of-the-art approaches to learning word embeddings are slow to train or lack theoretical grounding; We propose three spectral algorithms that overcome these limitations. All three algorithms harness the multi-view nature of text data i.e. the left and right context of each word, and share three characteristics: 1). They are fast to train and are scalable. 2). They have strong theoretical properties. 3). They can induce context-specific embeddings i.e. different embedding for “river bank” or “Bank of America”. They also have lower sample complexity and hence higher statistical power for rare words. We provide theory which establishes relationships between these algorithms and optimality criteria for the estimates they provide. We also perform thorough qualitative and quantitative evaluation of Eigenwords and demonstrate their superior performance over state-of-the-art approaches. Next, we turn to the task of using spectral learning methods for brain imaging data. Methods like Sparse Principal Component Analysis (SPCA), Non-negative Matrix Factorization (NMF) and Independent Component Analysis (ICA) have been used to obtain state-of-the-art accuracies in a variety of problems in machine learning. However, their usage in brain imaging, though increasing, is limited by the fact that they are used as out-of-the-box techniques and are seldom tailored to the domain specific constraints and knowledge pertaining to medical imaging, which leads to difficulties in interpretation of results. In order to address the above shortcomings, we propose Eigenanatomy (EANAT), a general framework for sparse matrix factorization. Its goal is to statistically learn the boundaries of and connections between brain regions by weighing both the data and prior neuroanatomical knowledge. Although EANAT incorporates some neuroanatomical prior knowledge in the form of connectedness and smoothness constraints, it can still be difficult for clinicians to interpret the results in specific domains where network-specific hypotheses exist. We thus extend EANAT and present a novel framework for prior-constrained sparse decomposition of matrices derived from brain imaging data, called Prior Based Eigenanatomy (p-Eigen). We formulate our solution in terms of a prior-constrained ℓ1 penalized (sparse) principal component analysis. Experimental evaluation confirms that p-Eigen extracts biologically-relevant, patient-specific functional parcels and that it significantly aids classification of Mild Cognitive Impairment when compared to state-of-the-art competing approaches
Uterine torsion in bovines
No full text<p><strong><span>Introduction</span></strong></p>
<p><span>Uterine torsion is a major foe in the field of bovine obstetrics, presenting serious difficulties for both cattle and farmer who is entrusted for its care. This complex syndrome arises when the uterus rotates on its longitudinal axis, putting the dam and her unborn calf in danger of death. For uterine torsion to be successfully resolved and the health of the dam and fetus in cows to be preserved, it is essential to comprehend its etiology, clinical presentation, diagnosis, and management.</span></p>
Need of Online Workshops for Research Methodology: Perception study of Indian Ph.D. Research Scholars
No full textDesign, construction and evaluation of lagrangian sensor particles for the flow behavior determination in a 200 l and 15000 l bioreactor
No full textStirred Tank Reactors (STRs) are frequently employed in the bioprocessing industry to produce bioproducts employing bacterial, yeast, or mammalian cell lines. The productivity of these reactors may be impacted when they grow in size due to the presence of heterogeneous zones. Using the Lagrangian measurement approach with free-flowing sensors is one way to look at these heterogeneities. In this thesis, two Lagrangian Sensor Particles (LSPs) that can be made using off-the-shelf components and a shell that can be produced by computer numerical control (CNC) machining are designed and built to circumvent this problem and study the heterogeneities in STRs. One type of LSP was built around an Inertial Measurement Unit (IMU), and the second was built around a pressure sensor. They are built to have a density between 1000 and 1005 kg m^-3. These LSPs are then tested in a 200 L and 15000 L reactor. Particle Tracking Velocimetry (PTV) is performed on the LSP in the 200 L reactor. The distribution of acceleration, velocity, and axial position is then determined using the LSPs. The data obtained via Particle Tracking Velocimetry (PTV) is then compared with the data obtained from the LSPs. A steady increase in axial velocity is seen with increasing impeller speeds, according to data from pressure and IMU sensors. Thus, a modular Lagrangian Sensor Particle (LSP) is developed, which can help investigate heterogeneities in STRs by making the LSP platform widely accessible, providing consistent results for all types of sensors. Enhancing the sensor module, the microcontroller, and the data processing method can further expand these findings.Deutsche Forschungsgemeinschaft (DFG
