Project Title: Software Internalization Student Participants: Mustafa Hasham, Rick Hochstetler, Shiv Siddhant Kaul Faculty Advisor: Dr. Denise Byrnes Bell and Howell creates Electronic Parts Catalogues (EPC's) for many car manufacturers in the United States and Europe. These EPC's are used in auto part stores and dealerships to look up and order parts. Bell and Howell is now moving into a few Far East countries such as Korea and China. The problem with this new move is that the languages of these countries are not supported by any of Bell and Howell's current EPC applications. The AMRE team investigated ways to support internationalization of programs through the use of Java. They also had to consider issues of how to enter text in the form of ideographs which current keyboards do not support. The knowledge gained from this group's research will tie into the Next Generation project which is currently underway at Bell & Howell.

The Bell and Howell Team: Siddhant, Mustafa, and Rick

Project Title: Hanging Banner Stress Prediction Student Participants: Jon Drover, Divya Thadani, Stephen Boughton Faculty Advisor: Dr. John Ramsay

A Graph of Two -Variable Equations considered for the Model

METROMEDIA TECHNOLOGIES specializes in large scale printing, such as billboards and large banners. The banners, printed on vinyl coated fabric, are very large and supported at the edges using different types of fasteners, such as grommets or ratchet straps. A major concern for Metromedia is that attachment locations are such that wind created stress does not cause sign failure.

Metromedia had a model which was being used to determine the maximum distance between fasteners to avoid sign failure. This model was based on a number of simplifying assumptions which were causing the map for fastening locations to be too conservative. In addition, there was some lack of confidence in the model among users. The AMRE project was to develop a new model serving the same purpose but with fewer simplifying assumptions involved in determining the acting forces in order to produce a more accurate fastener location map.

After analyzing the existing model, the AMRE team developed a new, more mathematically sound model serving the same purpose. The new model, which takes the biaxial extensions of the banner into consideration, gives more accurate prediction of actual stress at attachment points and, due to the more careful analysis, provides greater confidence among the model users. The "stretched" banner was modeled with a computer graphics program and then with a mathematical function. The function matched collected stretch and stress data comfortably within accuracy specification given by Metromedia and was then used to create a stress map along the boundary. From this, an easy calculution presents the user with banner attachment locations.

Project Title: Cycle Time Prediction Student Participants: Indradeep Ghosh, Mark Schlabach, Lisa Carter Faculty Advisor: Dr. Jim Hartman Rubbermaid utilizes a system of plastic production that involves injecting molten plastic into a mold, cooling it, and then releasing it from the mold. Each step of this process takes time, and the time which is required to complete one step of the process varies based on a number of variables such as type of plastic, mold size, and number of injection points. In order to make the most profit from a given product, Rubbermaid must try to maximize the number of items it can produce in a work day. The Rubbermaid AMRE team completed an analysis of the cycle time required to produce a plastic product. (The cycle time is the time it takes for the mold to complete a cycle -- from the closing of the mold for an injection to the next closing.) The team developed methods based on linear regression and neural networks in order to provide Rubbermaid with a reliable tool for predicting future cycle times. This tool will be utilized by Rubbermaid in the planning stage of a new plastic item to help estimate the profitability of the product.

The Rubbermaid Team: Lisa, Indradeep, and Mark

Project Title: GIS Community Internet Pages Student Participants: Benjamin Adair, Anant Padmanabhban Faculty Advisor: Dr. Denise Byrnes Technology Geosystems Inc., or TGS Technologies, contracted with the AMRE program to develop a prototype web site at GISLink.com. This web site is meant to serve as an attractive and useful site for the Geographic Information Systems (GIS) community. The project was divided into three tasks. Task 1: Stimulate network traffic through the site by creating "killer" content. In order to do this, the site would host the latest news, classified ads, and a directory to links within the GIS community. TGS provided design requirements and recommendations for these pages. Task 2: Design and implement a prototype Map Store. The Map Store would allow TGS clients to interactively order maps through the web site. Clients could order a specific area from a given map and have it custom made. TGS provided technical information about digital file formats and algorithms used by the company to modify, create, and customize maps. Task 3: Design an interactive distributed learning center (Dislecsys) for GIS. The dislecsys would allow GIS community members, and interested web surfers to interactively order or create a course of GIS study using the proposed web site.

The TGS Technologies Team: Anant and Ben

---

Last Updated June, 2000   Devoki Dasgupta dasgupdd@acs.wooster.edu Dustin J. Welty dwelty@acs.wooster.edu