Dennis Smid, PhD
Dennis Smid, PhD, Director - Inventory and Systems, Union Pacific Railroad, Omaha, NE 68179, 402/271-3535
Jo-Ann Kane, Manager - EDI Purchasing Projects Consolidated Rail Corporation, Philadelphia, PA 19101, 215/209-5056
Presentation Objectives. The role of Purchasing must be taken in the context of Supply Chain Management. One facet of this role is the management of information exchanges with suppliers. This presentation explores the purpose of information standards in a successful supply chain, using the Rail Industry as a case study. It relates the Rail Industry's recent effort to develop a standard for the exchange of electronic parts catalogs.
Supply Chain Management. The term "supply chain" operates on two levels. On the first and most obvious level, the term describes the commercial processes of moving goods from the point of manufacture to the point of consumption. The use of the word "chain" emphasizes that this process connects the buyer and the seller; it also accentuates the notion that the process is separable into "links", each of which is essential to the satisfaction of the requirement. On the second level, the term "supply chain" describes the information flows between the buyer and the seller. The most obvious flows are those which are associated with the physical movement of goods (e.g., shipment notice or invoice), but the term also extends to flows not tied to a specific buy-sell transaction (e.g., specifications or requests for quote). Each information flow is a "link" in the chain.
"Supply Chain Management" is the organizational effort devoted to maintaining and improving the supply chain. The term emphasizes that the role of Purchasing extends 'upstream' to/through the supplier and 'downstream' to/through the internal consumer of the goods or services. Furthermore, it emphasizes that the facets of the process - that is, the links - must be managed collectively, not in isolation, to achieve optimum results.
The Role of Information In Supply Chain Management. The most obvious indicator of a successful supply chain is the efficient and effective flow of goods: high-quality product, minimal inventory, short cycle times, highly reliable and repeatable processes. At a minimum, the information flow of a successful supply chain will exhibit these same characteristics. More often than not, however, the efficient flow of information - up and down the chain - is much more than a by-product of a successful supply chain; it is the enabler which permits the success of the physical flow. In fact, accurate information can replace physical resources (e.g., inventory or premium transportation) in the chain.
The surging development of communications and information alternatives has generated unprecedented opportunities to increase the flow of information within the supply chain. The facsimile machine was introduced and fully deployed through American business within a period of three or four years. Voice messaging - with all of its drawbacks - is almost as ubiquitous. Millions of personal computers have been installed in the past decade, and new subscribers to the Internet are measured by the thousands each day. Rapid advances are being made in Automated Voice Response and Optical Character Recognition. Unfortunately, there is a big difference between data and information. Moving lots of data quickly is now easy, but the supply chain achieves no value if the recipient cannot interpret and take action on the data.
Within a given supply chain, the exchange of data can become meaningful in two ways. The simplest approach is to establish key personnel as contacts throughout the chain and formalize each contact's responsibility and expertise. The success of this approach relies on the selection of personnel and their uninterrupted availability. The obvious alternative is to 'codify' some of the exchanges into precise computer formats, and set rules as to when and how the data will be exchanged. Both of these approaches, however, tend to evolve into a solution which is unique to given buyer/seller, making it difficult to replicate the process. Moreover, sustaining several pair-wise relationships can become a significant on-going effort. Instead of enabling the performance of the supply chain, the flow of information can become a constant impediment.
National Information Standards. In the early 1980's it became apparent that the effective exchange of information between organizations required a widely recognized and utilized set of standards. Admittedly this recognition did not foresee the concept of Supply Chain Management, nor did it predict the new communications developments of the past fifteen years. Rather the impetus arose from the more mundane realization that the paper-based method of conducting business resulted in high - and largely unnecessary - transactional cost. And with this motivation, the American National Standards Institute sanctioned a committee - which was given the code name X12 - to develop standards for the exchange of business data in electronic format, which is more commonly referred to as EDI - Electronic Data Interchange. Initially just a small clique from a handful of US companies, X12 has grown to several hundred members from all facets of North American business. To date X12 has published standards for over 150 different information exchanges, and EDI has touched virtually every US organization.
Unfortunately, the exchange of information is more difficult than it looks, and the X12 standards were not a panacea. In order to accommodate the rich diversity of conventions and protocols exhibited across the widely different industries that make up North American business, X12's standards are very generic and flexible. For example, the X12 format for a Purchase Order is extremely robust; virtually any order from any industry can be mapped to the standard, perhaps in several ways. For the practioner this approach is both the beauty and the curse of EDI. The same flexibility which permits a sender of a business communication to transmit a wide array of information makes it difficult for the recipient to interpret the result with precision. Standards greatly enhanced the meaning of data exchanges within the supply chain, but - by themselves - did not eliminate the pre-arranged agreements between the buyer/seller which made each exchange unique - and hence expensive.
Industry Guidelines. As the X12 standards evolved in the early and mid-80's several industries recognized the inherent conflict between X12's generic approach to standards and the corresponding lack of precision which impeded implementation. For example, in May of 1984 the Rail Industry Forum within the NAPM commissioned the Subcommittee on Information Standards to encourage and facilitate the deployment of EDI within the industry. The Subcommittee's name emphasizes its wider role in the standardization of all information flow within the industry, and it has served the industry well as the concepts of Supply Chain Management have unfolded over the past decade. The Subcommittee has met regularly since then and has successfully identified the specific components of the X12 standards which are pertinent to the railroads and their suppliers. At times this amounted to little more than a simple agreement on structure of an electronic message; at other times it required that one or more railroads modify an existing practice to bring the entire industry into alignment. The final work product is a set of published guidelines which document the format and structure of exchanged data within the industry, thereby achieving the supply chain vision of precise, electronic exchange of information. The effectiveness of the Rail Industry program is evident in that EDI is an established practice, with coverage for purchase orders and invoices approaching 75% of the line items purchased.
Importance of Industry Efforts. The discussion of the previous paragraphs can be summarized as follows. A successful supply chain requires timely and accurate flow of information between the partners, which in turn requires standards to govern the information exchanges. In the case of buy/sell transactional data, ANSI's X12 Committee has partially solved the problem with its standards for EDI. However, within the Rail Industry it was the collaborative effort of the member roads that added the necessary structure to make the informational flows meaningful.
Several comments are appropriate about the Rail Industry's experience with EDI before extending the discussion to electronic parts catalogs. First, while EDI does not extend to all flows within the supply chain, the process of setting EDI standards has been a valuable experience. Second, the challenge of standards development is a constant compromise between flexibility and clarity. Third, the development of information standards turned out to be of more benefit to the suppliers than the purchasers. Because the number of railroads is relatively small, and because each road is relatively large compared to many of the industry suppliers, the EDI guidelines served to set the requirements for the industry. While the cost of developing pair-wise solutions with each railroad may have been uneconomical for some suppliers, a single implementation that worked for all roads has been a cost savings. Fourth, and finally, the mindset which motivates industry work aligns with the principles of Supply Chain Management. That is, not only are the costs of the supply chain shared between the buyer and the seller, but some of the costs can be reduced only by improving the supply chain of your competitor.
Electronic Parts Catalog - Background. By its nature a railroad is an asset intensive company. Unlike a trucking company or an airline, a railroad company must maintain its own right-of-way, and the annual costs of replacing or upgrading its track, bridges, and signals requires hundreds of millions of dollars for each Class I road. However, at almost an equivalent rate, railroads also invest in locomotives, freight cars, and track maintenance machinery. A single new locomotive can cost over $2 million, a freight car can cost $40,000, and a large piece of track equipment can cost over $100,000.
In general, each railroad self-maintains its own equipment, seeking to achieve a strategic advantage through high utilization of its assets. This arrangement puts a railroad in the business of running one or more facilities dedicated to preventive and failure maintenance of complex pieces of large equipment. In this respect, a railroad is in the same business as the airlines, the Department of Defense, or repair depots for construction equipment. Terms such as reliability, maintainability, and mission success are shared with all these businesses. Specifications, maintenance procedures, and maintenance schedules are all important components of a good maintenance program. However, the primary maintenance document for a large, complex piece of equipment is the "parts catalog". The parts catalog is usually provided by the Original Equipment Manufacturer (OEM) at the time a piece of equipment is manufactured. The precise format may vary, but it usually consists of a series of annotated isometric drawings showing how parts are assembled, with callouts to replaceable parts in an accompanying table or list. It is regularly used by the mechanic to understand how a component can or should be disassembled and to identify replacement parts.
Until very recently, parts catalogs were published; that is, the OEM developed drawings specifically for the parts catalogs, prepared parts tables as text, and assembled a 'proof' of the final catalog. The production of the parts catalog was an after-market service, and generally was the responsibility of a department called "Technical Publications". From the railroad side, the management and distribution of the parts catalog was a critical function, one which took on increasing complications if and when the equipment was modified during its life.
In recent years, the railroads have found that maintenance effectiveness increases as maintenance practices are improved and as better information is delivered to the mechanic. In particular, computerized maintenance management software is increasingly being deployed and enhanced, advising the mechanic of maintenance schedules, proper maintenance procedures, and past maintenance activities for each piece of equipment. Meanwhile, links to supporting functions - notably purchasing and warehousing - are being established to shorten the cycle time between recognition of need and delivery of a replacement part. All of this is consistent with the urgency to return a shopped asset to service.
Starting in the late 80's, high resolution computer monitors, powerful PC's, and sophisticated - but affordable - computer graphics packages began to change the world of parts catalogs for both the OEM's and the railroads. Two patterns emerged from the OEM side. First, the options and possibilities for producing the parts catalog virtually exploded as desktop publishing became a practical alternative. Second, OEM's more closely examined linkages back to their Engineering, Manufacturing, and MRP systems so that the preparation of the parts catalog became less an after-market activity, and more a by-product of production. On the railroad side, several software providers emerged with products to 'deliver' an electronic image of a parts catalog on a PC. Several advantages to the electronic medium were immediately obvious: (i) the administration and handling of volumes of catalogs could be largely eliminated, improving both the access to the information and the cost to provide it; (ii) the software could facilitate a more efficient navigation through the catalog with meaningful images, hot points, and word searches; and (iii) electronic links could be made between the information in the parts catalog and procedures or material ordering systems.
Electronic Parts Catalogs - The Need For Standards. By 1993 or 1994, the cumulative impact of the technological possibilities was beginning to be felt in the rail industry. OEM's were re-engineering their processes to produce parts catalogs, in some cases targeting a paper-based catalog, but in some cases targeting a proprietary electronic format. Meanwhile at least three railroads were exploring tools to deliver parts catalog information on a PC. Common to each of these efforts was the need to capture the parts catalog information from the hard copy; each tool included utilities to scan the drawings, re-enter the tables of part information, and create "hot spot" links between the drawing callouts and the parts. A moderate-sized catalog required up to 200 hours of preparation. In hindsight, it is now clear that the industry was "dis-integrated", and had already consumed extraordinary effort and resources in potentially redundant or even conflicting initiatives. At the time the issues were far less apparent. The tip of the iceberg was a chance inquiry by one railroad to one supplier on the possibility of getting a parts catalog in a format which could eliminate the scanning and re-keying of data into their delivery system. This inquiry raised the specter of proprietary exchange standards for each railroad/supplier combination, and the likelihood that such a situation would increase costs in the supply chain instead of reducing them.
Electronic Parts Catalogs - Rail Industry Task Team. In early June of 1994, the Subcommittee on Information Standards established a Task Team to address standards for the exchange of Electronic Parts Catalog information. Unlike its role in the development of EDI standards, the Subcommittee possessed neither the organizational sanction, the expertise, nor a representative viewpoint from the supply community to develop an industry standard. Nonetheless, with the support of the Executive Committee and active proselytization from the Subcommittee members, the Task Team met in late June of 1994 with several roads in attendance along with two of the industry's largest OEM suppliers: EMD and GE. Perhaps more importantly, the representation from the railroads included maintenance and technical personnel in addition to representatives from Purchasing.
The continuity from the Subcommittee on Information standards provided two valuable leverages. First, while the majority of the task team members were unacquainted with each other, the core group from the Subcommittee served as a nucleus and facilitated the rapid assimilation of the other members. Second, while electronic parts catalogs are conceptually much different from EDI transactions, the process of setting standards turned out to be rather generic, and the "lessons learned" from a decade's experience with EDI were invaluable. Unfortunately, electronic parts catalogs differ from the transactional exchanges covered by EDI in two important ways. First, the content of the exchange is more complex by several orders of magnitude. Whereas EDI deals exclusively with data, the parts catalog deals with text, data, images, and linkages between these information types. Secondly, the Rail Industry approached EDI within the confines of the national standards set by the X12 Committee of ANSI. No such standards-setting body exists for parts catalogs.
Including its kick-off meeting in June, the Task Team met four times in 1994. Each meeting was a combination of cross-education and clarification of the Team's work product. The diversity of representation on the Task Team and the research undertaken singly built a good foundation - if not for solutions, then at least for an appreciation of the complexity of the problem. By the second meeting, the Task Team began to drill into the scope and requirements for the proposed standard. Three facets of the parts catalog were particularly troublesome. The first issue was the exchange of the isometric drawing. Taken to be an image (for example, as a scanned picture), the drawing was not complicated; several transfer and compression standards were available. However, isometric drawings are usually created from computer-aided drawing tools which construct arcs and lines from mathematical formulas. Such renditions are frequently called vector drawings. Converting a vector drawing to an image is easy, but all of the underlying intelligence is lost, making it very difficult to the recipient to make changes to the drawing. Since a typical locomotive has a life in excess of 20 years and since modifications to the original design are common, the railroads were most interested in receiving the drawing in its original vector format. Unfortunately, the standards and techniques for exchanging vector drawings are not yet mature. The second complication with the parts catalog was the issue of "hot spots", which is the ability to link a callout on a drawing to other drawings or to a particular location in a parts list. The concept is simple, and several software packages exploit this technique. The challenge was to develop standards which permitted the creator of the parts catalog to define these linkages. The third issue was how to package the parts catalog information into a single data exchange. Even if standards were developed for the individual pieces (text, data, and drawings), it was not clear how to organize these interrelated pieces so that the recipient could make sense of them.
As early as its second meeting, the Task Team recognized its limitations. Collectively the experience of the members was significant, but much too focused inside the industry. More importantly, none of the Team members was in a position to devote uninterrupted attention to the project, a fact which promised to unrealistically extend the duration of the project. The Team quickly decided to seek external support, preferably from professionals with connections with other industry groups, particularly with the Airline Transport Association (ATA), which has been very active in industry standards for maintenance information for several years. A short but specific list of requirements were drawn up, a list of fifteen potential service providers was developed, and requests for proposal were distributed. The Teams' third and fourth meetings included a short bidder's conference to review the request with potential respondents. The Team's activities created considerable interest within other standards' groups and among suppliers of proprietary parts catalog systems. Both bidder's conferences were well-attended and the Task Team members benefitted greatly from the discussions and perspectives of the attendees.
By bid closing on December 30, 1994 the Task Team received six proposals, each with a decidedly different approach to the problem. At the Task Team's fifth meeting the proposals were evaluated and consensus reached to further the negotiations with Applied Image Technologies (AIT) from Grand Rapids, Michigan. AIT is a software and service provider specializing in maintenance management tools, particularly tools to view parts catalogs. AIT's proposal included a sub-contract relationship with SoftQuad of Toronto, Canada, a small firm specializing in the use of SGML which is a structured format language. SoftQuad participates in numerous industry initiatives, and has strong ties with the ATA.
As the negotiations with AIT moved to clarify and finalize the contract language, the Task Team set about securing funding. The Rail Industry has Forum-status under the NAPM, and - as such - has a minimal balance of funds with which to support industry activities. However, it does not retain funds sufficient to support externally contracted work. With the support of the Executive Committee of the Rail Industry Forum, the Task Team solicited funds, not only from members of the Forum but from non-Forum companies represented on the Task Team. The solicitation took the form of an 'invoice' to the participating companies in an amount roughly commensurate with the company size. Response from the members has been extremely positive, and over $100,000 has been collected - more than enough to cover the AIT contract and associated costs and any expected publication costs.
The contract with AIT was finalized in April of 1995 and work commenced on the three-phase plan proposed by AIT. The first phase included a detailed report with two objectives: (i) a careful review of the external environment: other industry efforts and technological trends; and (ii) further clarification of the scope, objectives, and schedule for the remainder of the project. This report was accepted by the Task Team in early July. The second phase consisted of a systematic and comprehensive review of paper-based parts catalogs used in the Rail Industry, leading to further specification of the form and scope of the final standard. AIT reviewed its findings with the Task Team in early September of 1995, and was given the approval to proceed with the final phase of the project: the actual development of the technical standard. A draft of the standard is scheduled for release in late December with final acceptance of the standard expected during the first quarter of 1996.
Electronic Parts Catalog - Lessons Learned. While the ultimate success of the standards for the electronic parts catalog can only be judged as actual exchanges occur, the process of developing the standards appears to have been solid. Four points are worthy of note. First, the Rail Industry initiated the activity at approximately the right time. Enough activity was underway that sufficient interest was generated within the industry, but the development was not so far advanced that positions had become solidified. Second, the collaborative relationship within an industry benefits all participants, and drives down the overall cost for the industry. In this case, the purchasers of the equipment are setting a consistent expectation for documentation for the suppliers. Not only will this reduce the costs to exchange documents for a given buyer/seller, but it reduces the costs overall because each seller has only one target format and each buyer can be assured of inputs which meet expectations. Third, the Task Team's relationship with AIT demonstrates a powerful alternative to traditional committee work. The role of the Task Team shifted from 'doer' to 'reviewer', and the 'doing' was delegated to specialized experts. Finally, the role of Purchasing within the Rail Industry was advanced, consistent with the philosophy of Supply Chain Management. Without the Task Team's intervention, the evolution of electronic parts catalogs would have been driven by the maintenance professionals within the railroads in isolation of other considerations with the supplier. The stamp of the Purchasing group assures that the electronic parts catalog is properly positioned as another link in the supply chain.