Improving Revenue Forecasting Through Analysis of Historical Data
Team members: Robin Morillo, Ke Song, Pratistha Bhandari, Chloe Acheampong
Faculty advisors: Drew Pasteur (Math) and Harry Michael (Economics)
The AMRE team was hired by ArtiFlex Manufacturing to improve their revenue forecasting model. ArtiFlex manufactures replacement parts for automobiles. Throughout the eight-week long program, the students analyzed production and shipment records of ArtiFlex in comparison to their forecasting model predictions. Using Excel and other computational tools, the team was able to improve many areas of the forecasting model, improving its accuracy in predicting future revenue for ArtiFlex. In addition, the team updated the user interface of the forecasting model by adding features that made viewing and editing the data easier and more efficient.
Market Research and Business PlanDevelopment for a New Business
Team members: Robert Beall, Jack Marousek, Liam Fukushima
Faculty advisors: Brian Karazsia (Psychology) and Vikki Briggs
Traditionally, AMRE projects have been based in math and computer science; however, opportunities focused on entrepreneurship have been explored recently. Such is the case with Brant and Cochran, an early-stage premium axe manufacturing firm. Most research was delivered in a modified business plan. This consisted of conventional sections such as “Products and Services” and “Marketing Research and Plan,” as well as a plan for the coming twelve months. In addition, our team created financial documents and projections for Brant and Cochran, including a four-year pro forma income statement outlining multiple forms the company might take, along with several scenarios within those forms. Competitor’s sales and income were also analyzed to estimate the total market for premium axes (as measured by sales revenue and number of axes sold). We also laid the groundwork for the company to begin holding focus groups to finalize core user profiles, a defined product, and brand identity. Our team had a great experience being so exposed to the inner workings of an early-stage startup, local businesses, and the successful individuals who own and operate them.
Marketing Entrepreneurship and 3D Printer Market Research
Team members: Marla Walton, Alex Lalonde, Jacob Sparks
Faculty advisors: Peter Abramo (Center for Entrepreneurship) and Kate Gullatta (Learning Center)
The C4E AMRE team assisted the Director of Entrepreneurship in web page design and other marketing aspects of the Entrepreneurship program at The College of Wooster. The team researched entrepreneurship programs at other institutions and researched the feasibility of incorporating potential new initiatives at The College of Wooster. The main initiative that was investigated was a 3D printing program that would train students to act as consultants to business and industrial clients with 3D printing needs. The team developed a plan for the acquisition and implementation of a 3D printer and helped to shape potential program ideas for the 3D printer. The team was able to supply advice on what 3D printers would be of best use for The College of Wooster students. In addition, the team was able to redesign The Center for Entrepreneurship’s website to make it more informative and user friendly for both prospective students and current students.
Goodyear Tire & Rubber Company X-Ray Project
Team members: Emily Howerton, Alex Iudice, Khoa Nguyen, Obed Kobina Nsiah
Faculty advisors: Denise Byrnes (Computer Science) and Matt Moynihan (Math)
The purpose of this project was to provide Goodyear with a software tool to automate X-Ray based quality checks. The software is used to assist operators who observe X-Ray images of large tires for defects that can cause tire failure. Generally performing two major types of analysis (Splice and Waviness) the program generates results that the operators can use to make production decisions. For each type of analysis, a unique filtering method is used to process the X-Ray images into viable formats which make it easier for the computer to perform the analysis. At the end of the project, both analysis were finely integrated into an intuitive user interface to make user interactions with the software easy and interactive.
Tire Zigzag Belt Winding Optimization and Visualization
Team Members: Carlos Gonzalez, Kiera Dobbs, Joseph Smith
Faculty advisors: Drew Pasteur and Matt Moynihan (Math)
The purpose of this project wasto provide both tools and analysis in regards to optimal aviation tire wrapping. We began with a 2D model of the zigzag belt wrapping and weextended this into three dimensions in order to provide better visualization of the wrapping process. The finaltask was to create a design space to test possible wrapping patterns. This included using a mathematical approach to reduce the number of possible zigzag patterns to a small, optimal set. The result of this project was a 3Dprogram thatis equipped with a Graphical User Interface (GUI), allowing the user to input various tire parameters and then receive feedback of the resulting wrapping.
The Unknotting Problem
Team members: Brian Foley and Michael Bush
Faculty advisors: John Ramsay and Jen Bowen (Math)
The 2015 AMRE Knot Theory Research team worked on ‘the unknotting problem.’ The unknotting problem is an unsolved problem in knot theory that can be roughly described as follows: Given any mathematical knot – that is, a closed loop of string projected onto the plane with some number of crossings – the unknotting problem is to create an algorithm that determines whether or not that knot can be untangled to a circular loop with no crossings (the unknot), without cutting the string. Our approach to this problem was to design a set of generalized knot moves based on the Reidemeister moves, flypes, and contour moves. We designed an algorithm that intelligently applies these moves in an attempt to reduce the given knot diagram to the unknot. In addition to working on the algorithm, we also worked towards writing a computer program that implements our algorithm to untangle knots using numerical methods as opposed to pencil-and-paper application of our algorithm. Another key component of our work this summer was the development of a mathematical proof that our algorithm is a necessary and sufficient solution to the unknotting problem.
Tomato Analyzer 4.0
Team members: Nanako Ito, Nan Jiang, Max Taylor, Lydia Kemuma Kinyari
Faculty advisor: Simon Gray (Computer Science)
The OARDC AMRE team was asked to provide ongoing maintenance for Tomato Analyzer, a Windows-based C++ software application originally designed to perform analysis of images of sliced tomato fruits. Over the years Tomato Analyzer’s use has expanded to include seeds, leaves, and, most recently, peppers. The algorithms in the initial implementation of Tomato Analyzer assumed that the images to be analyzed were symmetric. However, this assumption has not extended to the kinds of images now being supplied to Tomato Analyzer. The majority of the changes requested dealt with improving the software’s ability to deal with asymmetric images. Additional functionality was implemented to deal specifically with peppers. The other task the AMRE team was given was to implement a version of Tomato Analyzer accessible through a web interface.
Building a Semantic Network for AAC Devices
Team members: Marissa Kobylas and Laith Sersain
Faculty advisor: Diane Uber (Spanish)
The AMRE team was hired to serve as consultants for Prentke Romich Company (PRC), a local company that makes augmentative and alternative communication (AAC) devices for individuals with disorders that prevent or hinder their use of spoken language. PRC is interested in making AAC software and language systems more intelligent by adding an automated recommendation feature that recommends related word sets to ultimately create a more organic learning process. The AMRE team was given the task of researching the applicability of a semantic network, a way to link words together based on their meanings, to AAC devices and developing a methodology for the creation of such a network. After conducting research and collaborating with PRC, the team recommended a hierarchically-structured database that links words together with several different Link types to indicate the type of relation between two given words. The recommendations for the system will allow for both flexibility and growth over time, and for a powerful framework that can adapt to users’ needs.
Data Analysis for Competitive Intelligence
Team members: Kenneth Mintah, Katarina Kremling, Sophia Anderson, Melissa Griffith
Faculty advisors: John Ramsay and Ronda Kirsch (Math)
The AMRE team had two projects: determining competitiveness in certain markets and analyzing a competitor’s tiering model. They were given data from various insurance raters to analyze characteristics of policy holders as well as how well Progressive’s premium ranked against other companies. They were also given competitor data to read, summarize, and apply to potential customer profiles. The team detailed their findings in a report to Progressive.
Standard Rate Adjustment Template
Team members: Josh Houtz, Gina Lam, Heather Smith
Faculty advisors: Jen Bowen and Ronda Kirsch (Math)
The Western Reserve Group was provided with five Excel files and one Access file including all of the queries, macros, modules, and formulas required to complete the tasks they assigned to the team. These documents can to be used in the future as templates for other applications. The AMRE Team’s hope is that the work they have done will be applicable to future projects in other lines of business, including farm and auto insurance.