Atmel is a large semiconductor manufacturer with a worldwide distribution of products. Engineers and management is having problems diagnosing problems with six machines that create semiconductors for the photolithography group at the company. Prior to this project, engineers were simply reading data from a database and making realizations on possible issues with no set data analysis to determining a problem. With this, there were many issues with human errors in reading the data since trends were hard to find. Our team was asked to create a graphical user interface for the data stored in the database so that engineers and management can make a more informed decision on where in the photolithography process improvements can be made.
Our team will be given a sample data set is asked to create the graphical user interface in Visual Basic .net and have it connect to MSSQL server. Main goals for this project is to allow management and engineers to be able to make informed decisions with little human data processing by showing machine and recipe trends through charts and diagrams; therefore, limiting the human element of data processing to a minimum.
Mr. Dusto also requested the website be prepared to add his chemistry honors course. He can then add his notes and use the site for the next school year.
Based on the host server and software we will add a database system
for grading quizzes which will require logins and passwords for each
student. If time allows we will create a fun new game to help engage
the students in the material. Improving the previously created web site
will enhance the website’s accessibility and create a better resource
for Mr. Dusto and his students.
Comverse develops and deploys world class billing and ordering applications for telecommunications companies worldwide. The Kenan FX line of products meet 60-80% of each clients' needs for customer services calls to service activation and billing. Custom software for unique client functionality is often written by consultants in the professional services division to fill the other 20-40%. The process of starting a new custom project is quite tedious. A consultant often starts with a basic template that gets modified before beginning the customization. These modifications include renaming files and directories, modifying contents of existing files, and copying directories and files. During the modification, if a single error is made it could be very hard to find where the error is. Every custom application for a client begins with this process; there must be a better way.
Our task is to develop an engine that reads in an XML file detailing changes to the template project, executes those changes, and returns the source for a new custom project. The engine would optimally be generic and flexible so that it can be extended for use in multiple settings. Time permitting a GUI will be developed that takes the information from the user and updates a template XML file to be used by the engine.
Part 1. Nonlinear Scattering of Bose-Einstein Condensates on a Finite Barrier
We seek stationary state solutions to the nonlinear Schrodinger equation (NLS) for piecewise-constant potentials V(x). The NLS is integrable on the infinite line, which means that it is exactly solvable in the space/time domain. However, this is not true for general piecewise constant potentials. In this case, we are able to find complete stationary solutions, i.e. exact solutions which are separable in time and space, in terms of Jacobi elliptic functions. These special functions can be defined as an inverse integral and are an interpolation between the circular and hyperbolic trigonometric functions.
As an important mathematical application of our exact solutions, we will seek a nonlinear generalization of the transfer matrix. When this matrix, which is commonly used in electromagnetic transmission problems, is applied to the coefficient vector characterizing the solution in the region to the left of the first boundary, it yields the coefficient vector for the Nth region. To solve this general problem, we begin with the nontrivial case of two boundaries, which corresponds to a potential step or barrier. This has a direct physical application to an atom laser incident on such a potential, as occurs in Bose-Einstein condensates on a chip. Our work provides the first step towards formulating a completely general nonlinear scattering theory.
Part 2. Numerical analysis of Schrodinger's equations
Bose-Einstein Condensates are a recent empirical discovery in the physics world. They represent a rich field of study because of their simultaneous quantum and macroscopic natures. They allow physicist a window into the world of quantum mechanics, a view that until now has been difficult to get.
A Bose-Nova is a Bose-Einstein Condensate trapped in a magnetic field, in which the magnetic field is tuned so the particles are mutually attractive. This causes the Bose-Einstein Condensate to collapse. This experiment has been performed at the JILA/NIST laboratory in Boulder, Colorado. What is interesting about this experiment is the conclusion of the collapse. Instead of simply becoming smaller, as experimentalists expected, the Condensate explodes, much like a miniature super-nova of the cosmological variety. Experimentalists named the phenomena because of the similarities. It provides a simple laboratory model of the phenomena of a super-nova as well, giving good reason to study it.
Theorists are still trying to understand why this phenomena occurs. They have theories, but have yet to demonstrate their theoretical results match up with experiment. In this project, we will be improving and rewriting a simulation of a collapsing Bose-Einstein Condensate, to show that our theoretical understanding of the phenomena matches experimental observations of the phenomena. We will improve a previous version of the code by introducing higher accuracy derivatives and rewriting it in C++ to increase readability.
After re-implementation, the code will be tested to prove its
stability, convergence, and norm error conservation. Throughout the
coding, refactoring will be done to ensure readability and
extensibility, and the project will conclude with a performance
analysis to ensure that the code is efficient.
For several years, the Colorado School of Mines has held workshops to improve the education of middle school teachers across the state. These workshops have been held in Colorado, instigated by the Mathematical and Computer Science Department. Due to the increasing desire to attend these workshops, we are providing an online version of the curriculum. Our team will be developing a website which includes each topic covered at the meetings and presents it in an online course manner. This will include the following:
There is a gross under-representation of minorities and females in the information technology and science, technology, engineering and math (STEM) fields. This gives need for hands-on activities where students can perform their own experiments to aid in the learning process. Information Technology Experiences for Students and T eachers (ITEST) proposes to meet this need by designing and implementing a year-round informal education program through the use of wireless sensor networks (WSNs).
The goals of the ITEST program are as follows:
We will use a collection of wireless, portable sensor nodes that
measure environmental and physical parameters, such as temperature and
barometric pressure, to create a wireless sensor network. Using this
network we will design activities that can be performed easily by
middle school students during allotted class times that will support
information being taught in normal class material.
Over the last five years, the number of students that have chosen to major in computer science has dropped by over 50%, nationwide. Thus, several educators have realized that younger children, specifically those in middle school, need to be encouraged to major in computer science. There is a website known as the Computer Club 4 Girls (http://www.cc4g.net/)that has many modules and activities to try and encourage young girls to develop an interest in computer science. However, this website was developed in the U.K. and may not be practical for the U.S.. Perhaps a similar website is needed for the U.S.. Thus, we have been tasked to:
The Los Alamos National Laboratory employees massive parallel computing systems to perform simulations ranging from biotechnology to nuclear weapons testing. As with any system running processes which could take days or even months to work through, system failure is an expected evil. While current check pointing practices and global, site-wide, disk storage have built a system of redundancy to protect against much information loss, failure handling protocols at LANL are far from perfect. In an effort to increase efficiency, LANL has released large data sets concerning these parallel processing systems ranging from usage data, physical position data, administrator-entered failure data, automated event log data, and node/domain usage data to the academic community. The sheer rarity of obtaining and analyzing such a data set is a huge reason for the analysis that we are pursuing. Also, it is highly likely that any predictive models, usage trends, or other findings which correlate to failure data could provide LANL with ideas for improved process and backup strategies. Our group is focused on evaluating possible correlation between these data sets using multivariate statistical methods and will be basing much of our work on the previous observations made by Bianca Schroeder and Garth Gibson from Carnegie Mellon University on only the provided failure data.
In order to ensure that our team is fulfilling all necessary project requirements, we have created a three-tiered objective list to organize our goals. Tier One objectives are the ones that we have decided are required to be met at all costs. Tier Two objectives are progressions of tier one objectives that we will address as time allows. Tier Three objectives are likewise an extension of Tier Two objectives but will only be addressed if there is significant time available. The three-tiered objective list is as follows:
Tier One
Los Alamos National Laboratory is a governmental nuclear weapons development facility. Additionally, the lab is a research facility for other scientific areas. They use large super computers to store their data and research. As data sets have grown increasingly large, the cost of searching the data has become progressively more expensive.
This project simulates large scale data searches by scanning multiple computers for JPEG image files. To accomplish this task, the project combines parallelism using MPI with efficient search techniques using Google Desktop and compares these results with those found using Windows' old file search. The efficiency of these data searches will be numerically analyzed and represented graphically.
Omnitech wants to have a report available each week (and on demand) displaying the status of the cash flows on all levels without the hassle of unnecessary data entry and other intermediate steps.
Our team will link a program to QuickBooks and will automate updates and calculations and make a report of cash flows available upon demand.
The goal of this project is to develop a library for virtual devices to utilize Simple Network Management Protocol (SNMP.) These devices exist within the Scalable Simulation Framework Network (SSFNet) and will monitor and store information related to their configuration and activity. We will further develop a MIB implementation which will provide network management data and an interface for updating any mutable entries. Our goals include: setting up methods to handle different data types that will be handled by SNMP, create a functional and well structured ipMIB, and use commercially available products, such as the Multi Router Traffic Grapher (MRTG), to demonstrate our implementation with graphical representations of critical SSFNet activity.
There will be two main databases that will provide the data, one from an accounting database and the other from a well activity database. However, the application should be modular enough to allow for simple database additions or be used for similar purposes in the future.
The language for this project will be C# on the .NET 2.0 framework.
Connectivity to the data will be via ODBC to Microsoft SQL Server 2005.
The goal of the project is to evaluate the performance of wireless sensor networks with varying parameters. The networks we will be working with will have no intelligent way of transmitting data, so our data will serve as a baseline for future tests of more intelligent protocols. The independent variables are network size, network density, and link quality. Network size is the number of nodes. Network density is related to how many other nodes each node can communicate with. Link quality is how well data is transmitted between two nodes in the network. We intend to measure how latency, packet delivery ratio, and energy consumption vary with each of the variables. Latency is how long it takes for data to get from a node to the base station. Packet delivery ratio is a measure of what percentage of the packets sent are received by the base station. Since transmitting takes more energy than computation, energy consumption can be estimated by how many messages are sent.
To accomplish our goal we are using TinyOS and its network
simulator, TOSSIM. In addition, we will use Python, C++, NesC, and Java
to varying extents. UNIX shell scripts should also be useful for
running a large number of tests easily.
Wendy Wood is a solo photographer specializing in various events such as weddings, family portraits, and graduations. Her current website is several years old and in need of a redesign and expansion of features.
The purpose of our project is to redesign the Wendy Wood Photography
website. The new website will require both a new design and a
reformatted, expanded structure. The new website should provide an
online shopping experience that will allow the user to view and select
proofs. This shopping experience should also walk the user through the
process of choosing picture sizes, photo quantities and other choices.
Finally, the new site needs to be easy to update for someone with
little programming knowledge. We hope to implement as many of these new
features as possible within the time provided.