York Laboratory Information Systems Restructuring Proposal IEM 5303

York Laboratory Information Systems Restructuring Proposal IEM 5303 Term Paper Fall 2000 Chris M. Forth Senior Project Engineer York International

Overview • • Background Problem Current Practices Alternatives Solution Evaluations Proposed Solution Future Considerations

York - Norman Operations • Commercial AC & heating manufacturer • UPG Division Headquarters • 539, 000 sq. ft. • 25, 000 Employees • Mfg. in 6 countries • Design Center • Laboratory test facilities • Since 1971

Background The York-Norman Laboratory is comprised of various individual test groups which generate or require the following types of data in the process of testing commercial air-conditioning equipment: • Psychometric Data (heating and cooling). • Heating Data (Gas and Electric). • Model Shop. Each of the above generate the following generic data types: • Live Data. • Historical Electronic Data. • Historical Hard-copy Data.

Problem The current laboratory information systems consist of multiple software products, located on multiple networks in various locations, in various formats, that prevents timely and accurate data accessibility, which ultimately lengthens the product design cycle for the Unitary Products Group of York International.

Current Psychometric Practices • Psychometric Group utilizes an outdated DOS-based data logging program (FIXDMACS), purchased in 1989. • Operates on a “stand-alone” DOS-based Novell network. • Can not be upgraded. • A large amount of historical hard-copy data. • No ability to support remote “live” viewing of data. • Limited data output saved as an exported *. prn file only. • A large amount of “non-convertible” electronic data. • Uses a Quattro Pro Interface for viewing only. • Considerable time & $ spent on maintenance of system.

Current Gas Heat Practices • • Currently utilizes a software product call Lab. VIEW. Purchased in 1998. Windows based, configurable. Data output saved as a *. xls and *. txt files. Resides on a dedicated Gas Heat network. A large amount of hard-copy legacy data. Not currently configured for “live” viewing of data.

Current Model Shop Practices • Requires information in the form of engineering part and assembly drawings (no numerical output data). • Engineer prints out hard copy drawing from CADKEY or Pro Engineer, physically delivers to the model shop. • Model Shop technicians fabricate sheetmetal and copper pieces, then assemble pieces into prototypes used for testing. • Model Shop archives the hard copy test data including drawings. • Limited training and / or hardware available to take advantage of electronic formats.

Solution Goals • • Integrate all data types (live, electronic, hard-copy). Minimize capital expenditures. Utilize existing, licensed software whenever possible. Offer the potential of remote data access. Utilize “plain vanilla” software whenever possible. Allow for future expandability. Must have export capability to MS Excel. Utilize in-house resources and skills wherever possible.

Live Data Alternatives • Replace the existing DOS-based FIXDMACS system with Lab. VIEW running on a Windows NT network for the psychometric rooms. • Upgrade the existing FIXDMACS system for the psychometric rooms as much as possible and continue utilizing the other systems as is. • Do nothing, leave the current systems as is.

Historical Electronic Data Alternatives • Utilize Lab. VIEW by programming an application that would first import the *. prn file and then export a *. xls file. • Write a custom Visual Basic program to do the same. • Do nothing and continue to import files as needed.

Historical Hard-copy Data Alternatives • Utilize “OCR” (optical character recognition) via scanning. • Create a MS Access database of the “test request” tracking numbers. • Hand key critical data into an MS Excel spreadsheet. • Do nothing, leave the current manual system as is.

Live Data Solution Evaluations • The “do nothing” alternative provides no integration benefits and consequently results in excessive cost due to the high maintenance requirements (~$45, 000/year) • Upgrading the existing system provides only marginal benefits due to the systems DOS-based requirements. • Lab. VIEW operates on a Windows NT network and offers an inexpensive client-based viewer for remote, “live” data access. It also has an established user base, site license, customizable, direct export capability to MS Excel which can be automated and “virtual instrumentation” which has the potential to significantly reduce hardware cost.

Historical Electronic Data Solution Evaluations • The “do nothing” alternative provides no integration benefits and consequently results in excessive cost due to the high maintenance requirements. • Writing a custom Visual Basic program would required inhouse program training or the expense of an outside consultant. • Writing a custom application within Lab. VIEW would allow for software standardization, utilize in-house resources experienced with the graphical “G” language and minimize cost by utilizing licensed software.

Historical Hard-copy Data Solution Evaluations • The “do nothing” alternative provides no integration benefits and consequently results in excessive cost due to the high maintenance requirements. • Utilizing OCR would require an estimated 139 days and ~$40, 000. • To hand-key selected data would require even longer than the OCR solution. • Creating a MS Access database of only the “test request” tracking numbers results in minimal cost, utilizes in-house resources / skills, allows for “ad-hoc” query capability as well as remote access when properly configured.

Live Data Proposed Solution • Due to the availability of a thin-client viewer and its existing in-house experience base, Lab. VIEW would be the best choice for viewing live data.

Historical Electronic Data Proposed Solution • Due to its existing user base and its existing in-house experience base, Lab. VIEW would be the best choice for the creation of a customized applications that would import the existing *. prn files and then export them as a formatted *. xls files.

Historical Hard-copy Proposed Solution • Due to its existing user base and experience, limited data input requirements and minimal cost, the creation of an MS Access database of “test request” tracking numbers would be the best choice integrating the historical hard-copy data. Note: • This solution would also require the manual reconfiguring of the existing hard-copy files in order to be effective.

Future Considerations • Investigate utilizing ISO 10303 (I. e. STEP) for all design and test data storage. This would allow for a standardized, vendor independent, long-term storage format that has the potential to further reduce the product design cycle. It is important to note that Pro Engineer currently offers STEP export capabilities for its geometric models that could be utilized now. It also has the ability to store more than geometric information via its product data management module (PDM).

Question / Comments? Send an email to: • Chris M. Forth • chris. forth@york. com (work) • cmforth@keytech. com (home)