The bill of materials (BoM) is, in its simplest form, a list of parts or components manufacturing has become increasingly distributed, the BoM has taken on even
25 pages

73 KB – 25 Pages

PAGE – 2 ============
White paper: The Perfect BoM Page 2 of 25 Table of Contents Executive Summary3 Problem Statement..4 Supporting Data8 Cost to the Industry..10 Solution Statement12 Education15 Recommendations.17 Glossary of Terms.18 References.19 Appendices20 A: Participating Companies B: Recommended Framework C: Sample PDX File

PAGE – 3 ============
White paper: The Perfect BoM Page 3 of 25 Executive Summary OEMs turn to outsourcing to ga in manufacturing efficiencies , but the efficiencies (and accompanying cost savings) they hope to achieve can be elusive. Some of the greatest obstacles are problems associated with bi lls of materials (BoMs). Supply chain interactions are seriously hindered by the lack of a standard data format for BoMs, unacc eptably high error rates and a lack of technology and tools to validate BoMs. The pr oblems are pervasive, and the consequences, far-reaching. The cycles required to clarify or co rrect data can directly affect time-to-market and time-to-volume. In 2001, the National Electronics Manufacturing Init iative (NEMI) organized a team of industry leaders (see Appendix A) to investigate and define the issu es surrounding BoMs. This team is recommending a solution that calls for industry to adopt a common format with standardized contents Š the fiPerfect BoM” Š to ensure that da ta is consistent, correct and complete, and that errors are resolved before exchanges are made among business partners. The team has identified the Product Data eXchange (PDX) specificati ons, developed by NEMI™s Virtual Factory Information Interchange Project, as the foundation for the Perfect BoM. These specifications provide an industry-standard a pproach to exchanging BoM and change order information and will be used to define data structure. In addi tion, the team has developed a recommended list of data to define standard BoM content. Implementation of the Perfect BoM will require a higher level of industry awareness and a commitment to standard solutions. The NEMI team™s action plan calls for an education effort to raise awareness of problems and proposed solutions in an effort to encourage rapid acceptance and implementation. This paper discusses the problems and challenges that exist with toda y™s BoMs, and outlines NEMI™s recommended solutions for crea ting and implementing the Perfect BoM.

PAGE – 4 ============
White paper: The Perfect BoM Page 4 of 25 Problem Statement The bill of materials (BoM) is, in its simplest form, a list of parts or components required to build a product. It provides the manufacturer™s part number (MPN) and the quantity needed of each component. At its most complex, the BoM is a multi-level document that provides build data for multiple sub-assemblies (products within products) and includes Š for each item Š part number, approved manufacturers list (AML), mech anical characteristics and a whole range of component descriptors. It may al so include attached reference files, such as part specifications, CAD files and schematics. Originally used internally within a company, the BoM served as a means of tracking product changes and maintaining an accurate list of components needed to build products. As manufacturing has become incr easingly distributed, the BoM has taken on even greater importance, serving as the primary reference file for product data. It is used to transfer product information from OEM to EMS provider and from the EMS to its vendors and suppliers. As outsourcing expands the number of companies involved in the manu facturing process, the need for accuracy is more critical than ever. At th e same time, the transfer of manufacturing data across multiple companies magnifies problems and increases challenges. As the primary conduit for data transfer among ma nufacturing partners, the BoM is central to the product life cycle from the very beginning. Figure 1 illustrates the flow of information from design to manufacturing. As indica ted, a diverse set of informati on systems feed into the BoM, each providing different pieces of the informati on required to manufacture the final product. Typically, this wide variety of design, PDM and ERP tools output BoMs in different formats and at varying levels of completeness. File t ype and structure vary by company and, sometimes, even within a company, particularly if seve ral groups have been merged into a single organization. This lack of standard format leads to redundant data processing. EMS providers use parsers to fineutralizefl data Š i.e., convert it to a neutral format, such as a standardized text or XML file, that can be used by the EMS provider™s tools. Parsers extract key data from customer data files and build a sta ndard data format. Today, EMS providers must

PAGE – 5 ============
White paper: The Perfect BoM Page 5 of 25 develop parsers and scripts for each different form at received. One EMS provider reports that it maintains an internal organization of about 80 people whose function is to interpret customer data and translate it into their own production system. Figure 1: This diagram represents the flow of information between OEM and EMS provider, from design through manufacturing. Data Errors / BoM Defects The most significant problem plaguing BoMs is e rrors. EMS providers report that the data received from their OEM customers is often inco mplete, inconsistent and/or outright incorrect. One EMS provider told the NEMI Perfect BoM team that seven out of 10 customers consistently need BoM corrections, and three in 10 compone nts have bad or missing data. Several EMS providers said that error rates of 80% are not uncommon. PDMSchematic Capture Layout PDM Mechanical Design BoM CreationValidationERPShop Floor Systems Item Master CAD Library item master flow design to manufacturing flow Component Selection EMSOEM

PAGE – 6 ============
White paper: The Perfect BoM Page 6 of 25 EMS providers are spending a great deal of time and resources co rrecting deficiencies in the BoMs they receive. Errors must be communicated back to the OEM for resolution, and valuable cycle time is lost. BoM errors typically fall wi thin three categories: completeness, consistency and correctness. Completeness . Incomplete data is th e most common BoM defect. Critical pieces of information, such as quantity, part description, reference designation and approved manufacturers list, are often omitted. Missing AMLs reportedly cause the majority of problems. Consistency. Information in the BoM sometimes c onflicts with information provided in engineering drawings and design files. For example, quantities may not match Š there may be 10 locations for a particular component indicat ed on a board, but only a quantity of nine components specified in the BoM. Another consis tency problem is format. The format of the BoM, even though it is from the same customer, can change from one transmission to the next, making it difficult to match and confirm data. Even language can vary from BoM to BoM. Correctness. Incorrect data is a serious problem. Common errors include invalid manufacturer or supplier information, obsolete data and incorr ect part numbers (i.e., the MPN given does not match the description of the part, or the MPN is not recognized by the ma nufacturer/supplier). Again, AMLs seem to be the predominant problem. Additional errors can result from receipt of information in hard copy format, which require s manual entry of date, an error-prone and time- consuming task. One leading EMS provider reports that, in the first two to three months of engagement with a new customer, an average of 80% of the BoMs received contain errors that must be resolved with the OEM. Of the component informati on received, 40% of the parts are found to have errors that require correction by the OEM. Thes e errors include inconsis tent content, invalid manufacturer or supplier information, incomplete or truncated information, missing information, obsolete data, and customer-specific information that requires the custom er to send a letter of authorization before the supplier will release information to the EMS.

PAGE – 8 ============
White paper: The Perfect BoM Page 8 of 25 Supporting Data The NEMI team gathered anecdotal informa tion from several OEMs and EMS providers, and discovered that companies consistently cited erro rs and lack of a standard format as leading causes of problems when working with BoMs. Es timates of error rates and of the time required to resolve data problems were also similar from company to company. Figure 2: This chart indicates defect levels experienced by an EMS provider at various steps in the BoM data set-up process. One major EMS provider provided estimates of def ect levels according to process steps. This information is depicted in Figure 2. Their estim ate of an 80% error rate for neutralizing and validating BoMs received from customers is surpri singly high but, unfortunately, does not appear to be unique, based on information received from other EMS providers. Data collected by a second EMS (see Table 1) corroborates and quantifies the problems identified by the first EMS. In particular, their data highlights the pervasiveness of incomplete and missing information. Neutralize & Validate BoM Validate & Set Up Components SourceMaterials BoMReleased Materials Pipeline EstablishedInconsistent format Hard copiesInconsistent with CAD file Foreign languageInconsistent contentsInvalid supplier info Incomplete or missing info Customer- centric info Allocations* Supplier withholds customer info Obsolete parts ISSUES––.. 80%40%20% *Allocations are important but beyond the scope of Perfect BoM content

PAGE – 9 ============
White paper: The Perfect BoM Page 9 of 25 Neutralize & Validate BoM Validate & Set up Components ( Yes / No ) ( # / total ) Customer Inconsistent Format Hard Copies Inconsistent with CAD File Foreign Language Inconsistent Contents Invalid Supplier Info Incomplete Info Missing Info Customer A NO NO NO NO 0/6 N/A 0/6 0/6 Customer B YES NO NO NO 0/1 0/1 0/1 1/1 Customer C YES NO NO NO 0/40 N/A 10/40 10/40 Customer D NO NO NO NO 0/6 N/A 6/6 6/6 Customer E NO YES NO NO 0/1 N/A 0/1 0/1 Customer F YES NO YES NO 0/30 25/30 25/30 25/30 Customer G YES NO YES NO 20/200 N/A 100/200 10/200 Customer H YES NO NO NO 0/200 50/200 150/200 10/200 Customer I YES YES NO NO 2/5 5/5 1/5 1/5 Customer J NO NO YES NO 0/10 N/A 0/10 0/10 Customer K YES NO NO NO 25/80 N/A 60/80 60/80 Customer L NO NO YES NO 0/30 N/A 2/30 2/30 Customer M NO NO NO NO N/A N/A N/A N/A Customer N NO NO YES NO 0/8 N/A 1/8 1/8 Customer O NO NO NO NO 0/8 0/8 2/8 2/8 Customer P NO NO NO NO N/A N/A N/A N/A Table 1: Defect data collected by an EMS provider. A leading OEM that works closely with its EMS providers to resolve data issues told us: Ability to start up new products rapidly is impacted by the quality of our documentation. Thirty-five percent of the component data have problems. If we do not have a process in place that will check and correct the errors in the audit report prior to sending the AVL to the EMS, these errors will continue to occur. Another leading OEM made the following obs ervations, based on a study it conducted to characterize internal processes involved wh en outsourcing to contract manufacturers: The number-one outsourcing issue identifie d was related to sharing data with EMS providers. There is no standardized or automated way to exchange data and, most of the time, the process is accomplished by manual means– There seems to be no clear ownership of the correct BoM version being transmitted

PAGE – 10 ============
White paper: The Perfect BoM Page 10 of 25 to the EMS, which means invalid data may be exchanged. Current perception is that the ERP BoM information is only about 80% accurate. The document control group of a major EMS provider had this to say: I agree with the estimate that 80% of the BoMs I get have some kind of a problem. They are almost al ways incomplete in terms of what we need to estimate product cost. We usually e nd up asking more questions of the customer, searching through all the other assembly files we receive with the quote package, or making an educated guess at the missing information. Some items that almost al ways need attention are: Assigning schematic reference numbers to hardware. Reconciling hardware requirements be tween the BoM and what’s actually required. Removing items that should not be part of the BoM (such as specifications, drawings, etc.). Structuring programmed devices in a pa rent/child relationship (so that, for example, a blank part and the software needed to program that part are differentiated). Cost to Industry Despite the pervasiveness of the problem and the magnitude of its impact, the issues associated with BoMs do not have a high level of visibility among senior managers in OEM and EMS provider organizations. Time and energy are spent on dealin g with the immediate day-to-day issues and fifire-fighting,fl while little or no time is spent addressing the root of the problem. The perception is that, as long as the EMS has receiv ed data, the situation is manageable. Little thought is given to whether the data sent is error-free, how much time is required to ficleanfl the data, and what the impact on cycle time is. The reality is that more re sources are devoted to correcting data than to correcting the problems that cause bad data in the first place Š i.e., to make sure that the data is clean and consistent be fore it is passed to a partner. The industry™s design, PDM and ERP tools are all creating different file formats. Time and energy must be spent on developing format translator s, or data neutralizers, before the data can be evaluated for accuracy and consistency. Currently, none of the leading PDM or ERP systems includes tools to validate data for accuracy and completeness.

PAGE – 11 ============
White paper: The Perfect BoM Page 11 of 25 The time lost in dealing with data issues can gr eatly impact the ability of the EMS to deliver products to its OEM customers on schedule. Each de fective part adds an estimated 40 minutes to the processing time of the BoM, and the lag time is much longer. On average, delays can range from one day to several weeks, depending on th e responsiveness of the information provider. For example, if a BoM has 100 unique parts and 40 of the parts have issues, the processing time would increase by 26 hours or approximatel y 3.5 working days (assuming the EMS and OEM resolve the discrepancies right away). If the queue/wait time required for resolving issues is added to the calculation, the overall cycle time can increase by weeks. One leading EMS provider estimate s that it takes 10 minutes of e ngineering time to analyze each error and recommend a correction to the OEM. When the volume of parts produced and defect rate are factored in, the time spent on correcting errors beco mes significant. For example, 10,000 parts (which is what a leading EMS typically sets up on a quarterly basis) with a 40% defect rate would require 40,000 mi nutes (66 hours) of engineering time. In addition, there is time spent tracking each issue, reporting back to the OEM, following up and finally entering the correction into the systems. Another EMS reports receiving a 75-line-item BoM with 40 errors. It took two days of cycle time to prepare the product for manufacturing. Yet another EMS provider estimates that, given the volume of BoMs and new parts going throu gh their system, they are losing six person- months each month to finding, reporting and correcting data errors and inconsistencies. The greatest cost, which is difficult to quantify, is how lost time impacts time to volume and time to market. Until the last issue on a BoM ha s been rectified, the product build cannot be completed and, in many cases, the manuf acturing process cannot even begin.

73 KB – 25 Pages