Avoid Million-Dollar Manufacturing Disasters

In engineering, something as simple as declaring the wrong unit of measure can create a disaster costing millions of dollars. John Pike, space policy director at the Federation of American Scientists, stated, “It is very difficult for me to imagine how such a fundamental, basic discrepancy could have remained in the system for so long,” regarding the English to Metric units mistake at NASA [1].

Watch the video overview. Article continues below.

Four Threats to On-Time and On-Budget

Engineering disasters can often be traced back to design errors that creep into downstream manufacturing processes through computer-aided design (CAD) software processes such as poor revision control or data compatibility issues. Even small mistakes might have big repercussions, valued in millions of dollars, from lost revenue, corrective work, time delays or investment loss.

Reviewing past engineering disasters provides critical information for avoiding expensive and high-profile mistakes by creating engineering processes that ensure accuracy throughout the design to production lifecycle. Author of To Engineer is Human, The Role of Failure in Successful Design, Henry Petroski states, “…the colossal disasters that do occur are ultimately failures of design, but the lessons learned from those disasters can do more to advance engineering knowledge than all the successful machines and structures in the world” [2].

Classic Disasters

French Regional Express Train

Cost of Mistake: $68 Million

Paris, France – May 21, 2014 – France’s main train company, Réseau Ferré de France (RFF), spent approximately $1.83 billion dollars on a new fleet of trains. When preparing to unveil the new fleet, RFF realized that the new trains were too wide for “hundreds of stations.” The station dimensions given to Société Nationale des Chemins de fer Français (SNCF) by RFF were taken from the newer train stations, neglecting to take into account older stations with narrow design. As a result, the error will cost another $61 million dollars to modify the older stations to accomodate the new trains [3].

NASA Genesis Probe

Cost of Mistake: + $260 million

Utah, United States – September 8, 2004 – Launched on August 8, 2001, the Genesis probe was sent to space to collect samples and solar wind data. On September 8, 2004, the Genesis probe crashed to earth, in Utah. The crash was caused by a critical error: a pair of the probe’s accelerometers were installed backwards. The inversion ultimately prevented the probe from properly deploying its parachute for a landing. The estimated cost of the mistake is over $260 million, plus the loss of data since the crash-landing contaminated many of the probe’s samples [4].

The Hubble Telescope

Cost of Mistake: $1.5 billion

How big is 2 microns
Figure 1 – How big is two microns?

Space – December 2, 1993 – Launched in 1990, the Hubble Telescope orbits around Earth, relaying back imagery. Shortly after launch, it was found, to the disappointment of many, that the images sent from the telescope were much lower quality than anticipated. What caused that?! The telescope’s mirror had been ground down 2 microns (0.00007874 inches) too thin (see Figure 1). The result was sub-optimal light direction generating poor quality images. A repair crew was sent into space on December 2, 1993 to correct the issue using additional mirrors designed to compensate for the original mistake. The cost of the mistake was $1.5 billion [5].

Mars Climate Orbiter

Cost of Mistake: $328 million

Mars – September 23, 1999 – Launched December 11, 1998, the Mars Climate Orbiter’s intended mission was to collect data on the Martian climate. The Mars Climate Orbiter, expecting navigation data in metric units, was sent data in English units [6]. The result was the $328-million [7] Mars Climate Orbiter ventured too far into the Martian atmosphere, where it subsequently overheated and disintegrated [8]. Noted in the NASA report, “‘The ‘root cause’ of the loss of the spacecraft was the failed translation of English units into metric units…’” [9].

Airbus Wiring Harnesses

Cost of Mistake: $6 billion

Hamburg, Germany – October 4, 2006 – A recent mismatch occurred with the Airbus 380 due to “design software used at different Airbus factories (not being) …compatible” [10]. The electronics harness designers in Hamburg, Germany, had an older version of CATIA, a 3D CAD Software, than the assembly plant in Toulouse, France [11]. Although both had CAD software from the same company, data conversion between the mismatched software versions resulted in hundreds of miles of wiring harnesses that did not fit in the planes waiting in Toulouse. What was the result?! A 2-year halt in production, with over $6 billion dollars in losses and changes in upper management [12].

The engineering mistakes that come with large price tags often make the news; however, the reality is that engineering disasters can strike anywhere. From bridges to buildings, airplanes to trains, engineering and manufacturing industries create today’s world. One bad file translation, one overlooked design revision or one unchecked design can derail a production schedule and potentially cause irreparable harm.

Human error and design flaws are commonly cited as top causes of past engineering disasters [13]. In analyzing past disasters for lessons learned, communication failure [14] and inadequate checks and balances [15] are top reasons these issues exist. That is, poor communication and inadequate processes to ensure proper validation are the core origin behind human error and design flaws, sometimes leading to engineering disasters. Noted in the Vanderbilt University study, “The inability to share important information that is timely and accurate is a common denominator in every case we reviewed” [16]. In conjunction with poor communication, processes “…involved a lack of checks in the design and implementation of whatever failed” [17].

AVOID THESE FOUR THREATS

Avoiding future engineering disasters is a crucial objective. Learning from the past to create better processes today can not only avoid the more serious million dollar mistakes but can also aid in maintaining production schedules, budgets and success rates while minimizing waste and lost revenue. Pulling from these case studies, Four Threats to Manufacturing On-Time and On-Budget are:

  1. Data Lost in Translation
  2. Excessive and Unchecked Changes
  3. Wrong Parts and Missed Deadlines
  4. After-the-Fact Errors

Threat 1: Data Lost in Translation

As technology continues to permeate everyday life, “computer-aided (CA) technologies play an increasingly important role in modern industrial practices” [18]. In addition to “neutral” CAD file formats, each computer aided technology (analysis, machining, simulation) includes its own list of proprietary data file formats. Prominently noted, “Integration and interoperability issues are still unsolved” [19]. Moreover, 75% of large design issues are attributed to data exchange, “directly related to differing CAD versions or systems, file formats and conversions carried out” [20]. Design data exchange is commonplace with 42% of users exchanging 100 or more files a month – 5 designs each business day [21]. Compounding the CAD data exchange issues, poor quality file translations result in enormous time and energy spent reworking design data [22]. Approximately 1 in 5 imported CAD models still contain errors that need to be manually fixed before they are usable [23]. In a week, about 1 out of 7 engineers spend over 24 hours fixing design data [24]. Approximately 49% of engineers spend at least half a day per week repairing design data; half of engineers work overtime or on the weekend reworking geometry [25].

Data translation and geometry reworking creates potential for data loss and errors. In turn, data loss and errors can result in a series of consequences. While the least serious repercussions from data loss could be re-tooling or project delays, more serious consequences also occur, costing millions of dollars, and, in some cases, irrecoverable damage.

“Two manufacturers cannot collaborate if the systems in one cannot process the data in the other” [26]. The Mars Climate Orbiter was “lost in translation,” [27] literally describing its disintegration

in space from the Metric-English data loss. Data loss cost Airbus $6 billion dollars due to CAD software version discrepancies [28]. Within the multiple CAD software environment, design data exchange is critical as “…product models can only be applied effectively if data can be exchanged and/or shared freely…” [29].

Threat 2: Excessive and Unchecked Changes

Regional Express Train didn’t think twice about the station dimensions they were given by SNCF; the cost of failing to validate their data was $68 million, the additional cost to update the older stations [30]. The Hubble Telescope was 2 microns out of tolerance; failing to adequately compare the manufactured part to the original design cost $1.6 billion [31]. In the midst of enormous pressure to get products out on schedule, opportunities for error increase due to excessive and unchecked changes and data.

Approximately 44% of CAD users experience excessive, unanticipated changes throughout the product development cycle [32]. Almost 38% of users experienced dramatic or multiple last-minute changes to product definition [33]. Even in a high-profile expensive project like the Mars Climate Orbiter, engineers failed to check mathematic basics, units of measure [34]. High-pressure, last-minute and unchecked changes could exponentially add to wasted material, scrap material and valuable time loss. In the worst case scenario, unchecked changes and data results in project failure.

Threat 3: Wrong Parts and Missed Deadlines

While CAD files may be translated accurately and verified, additional product design information may still be lost. Creating a significant time delay, the process of “understanding of the data can be delayed for weeks or months” [35]. Importantly, missing a CAD design change, or recording the wrong dimensional value can cause an entire project to miss delivery dates. While the Genesis Probe had the right parts, they were installed backwards, causing them to function incorrectly and destroy a project costing $260 million dollars. “Errors that occur in the exchange of product data are caused by differences between the standard and the internal schemas of the CA-applications” [36]. The Mars Climate Orbiter project failed due to a miscommunication on the units of measure!

About 32% of organizations miss project deadlines, and 29% have ordered incorrect parts due to design data problems, poor tracking of design change orders and downstream impacts [37]. The impact of imperfect design data flow affects the entire supply chain. The US Automotive industry estimated that $1 billion dollars was lost each year from CAD data interoperability [38].

The main issues facing engineering and manufacturing sectors is the struggle regarding “automation, information management and computerized design and planning techniques” [39]. It is not enough to have an accurately translated model; users require complete product manufacturing information in order to appropriately contextualize the design within its intended use.

It is notable that, “in manufacturing, the multifaceted nature of design information makes communications particularly difficult. A mechanical design may contain geometric, tolerance, material, process control, and many other kinds of information” [40].

Threat 4: After-the-Fact Errors

As in many situations, the sooner errors can be caught, the better the ultimate result. For the Hubble Telescope, the error in the mirror, ground 2 microns too thin, was not caught until the telescope was already orbiting Earth, resulting in a $1.5 billion dollar space repair mission. Catching errors and mistakes after the fact often results in higher ultimate costs from wasted materials, time and production capacity. In the automotive industry, erroneous CAD geometry was not discovered until part tooling had already been designed and cut in 15% of the cases [42]. Design mistakes, revisions and missed errors have downstream ripple-affects, with the ability to impact tooling, mold development, fixture design, machining, even assembly and packaging.

See the Cost of Change paradigm shown on http://www.agilemodeling.com/essays/costOfChange.htm

The Solution: Accurate 3D CAD Design Comparison and Validation Software

In the NASA board report on the Mars Climate Orbiter disaster, one of the eight contributing factors was “the systems engineering function… that is supposed to track and double-check all interconnected aspects of the mission was not robust enough…”[43]. Uncovered in various independent analyses of past engineering disasters, communication and data validation are critical factors for avoiding human error and design flaws; engineering design checking, comparison and validation are paramount to avoiding million dollar mistakes. Whether verifying two design revisions for changes, comparing the manufactured part to the CAD model or validating a clean file translation, checking and double-checking engineering designs is crucial. Just as important, the ability to seamlessly communicate that information downstream and across the organization avoids the miscommunication and lack of information pitfalls in manufacturing. The solution is two-part: Design Validation and Information Communication.

To avoid million dollar mistakes, the solution must possess the ability to compare, validate and verify 3D CAD data. The 3D CAD comparison process should be clear, accurate and easy, in order to maximize compliance and minimize wasted resources. With information communication being the second critical juncture, results should be easily sent downstream with the model.

Figure 1 - Sectioned view of Revision Analysis.
Figure 2 – Sectioned view of Revision Analysis.

Comparison software can provide visual evidence of changes to a model (Figure 2).

If the two models are overlaid, the differences are visually apparent, making a quick assessment possible; this is superior to having to look at two different screen images and guessing what has changed. Both models can also be shown separately if desired. Dynamically slicing the model through the X, Y and Z planes displays changes inside the geometry that might otherwise be overlooked or hidden from view.

With controllable tolerance variations, the software will highlight even the smallest change between 3D design files; any unintentional or accidental geometry changes in the design part will be numerically and graphically displayed. A visual color plot shows all changes such as if the holes moved or have new chamfers. Additional edits to model geometry introduced in the manufacturing processes will also be highlighted; validation of changes is an easy check with a CAD comparison solution.

Accurate CAD Comparison in 3 Easy Steps

  1. Set the tolerance level for model deviations
  2. Open design files and use “Compare” feature
  3. Show changes, revisions and deviations
revision-analysis-process
Figure 3 – The Revision Analysis Process can also be used for Validation Analysis.

As Figure 3 shows, generating a heat map of areas where the part is out of tolerance as compared to the original is a three-step process. You open both parts up, set the tolerance, and click ‘Analyze Parts’. This process is the same whether you are comparing revision A to revision B, or comparing an original CATPart model to a derived STEP model for validation analysis. For Revision Analysis, just click the Revision Analysis radio button. For Validation Analysis, just activate that button.

Revision Analysis Report
Figure 4 – MagicCheck allows you to easily generate a Revision Analysis or Validation Analysis report.

To aid in the flow of information between departments and suppliers, the ability to easily generate a comparison report is critical. A default report is shown in Figure 4. Report generation from the true CAD model inherently delivers higher accuracy results and less opportunity for human error. The report includes BREP model visualizations, accompanying model data and analysis specifications often required downstream.

With 3D CAD comparison software based on the actual geometric data, engineering designs and parts can be checked for file translation validity and design changes can be verified; manufactured parts can be validated to the original design and serve to catch any errors prior to production. Robust CAD checking and double-checking processes increase the fidelity of the design to manufacturing process. Using CAD comparison software can have an immensely positive effect on manufacturing schedules and budgets, knowing with certainty what changed, where it changed and how much it changed.

Using computing power to detect small or hidden areas of discrepancy surpasses the old and inefficient methods of manual visual inspection, removing the potential for human errors in inspection. Empowering users with the ability to find and solve issues quickly generates a competitive advantage, confidence in 3D CAD geometry, verification prior to production and, most importantly, reduces the risk of engineering mistakes and million dollar errors.

Try CAD Revision Analysis for Yourself

CAD Revision Analysis and CAD Validation Analysis are both part of the MagicCheck add-on, which is included with every evaluation version of TransMagic.

References:

1 Hardwick, Martin, David L. Spooner, Tom Rando, and K. C. Morris. “Sharing Manufacturing Information in Virtual Enterprises.” Communications of the ACM 39.2 (1996): 46-54. Communications of the ACM. Communications of the ACM, Feb. 1996. Web. 24 Dec. 2014. http://diyhpl.us/~bryan/irc/step/alvarestech.com/temp/0steptools/p46-hardwick-empresavirtual-stepnc.pdf.

2 Gielingh, Wim. “An Assessment of the Current State of Product Data Technologies.” Computer-Aided Design 40.7 (2008): 750-59. Web.

3 Neild, Barry. “French Trains ‘too Wide’ to Fit in Some Stations.” CNN. Cable News Network, 21 May 2014. Web. 23 Dec. 2014. http://www.cnn.com/2014/05/21/travel/france-trains-sncf/index.html.

4 “Six Tiny Scientific Mistakes That Created Huge Disasters – World Science Festival.” World Science Festival. N.p., 04 Nov. 2014. Web. 23 Dec. 2014. http://www.worldsciencefestival.com/2014/11/six-tiny-scientific-mistakes-created-huge-disasters/.

5 “Hubble Essentials.” HubbleSite. N.p., n.d. Web. 22 Dec. 2014. http://hubblesite.org/the_telescope/hubble_essentials/.

6 Isbell, Douglas, and Don Savage. “MARS CLIMATE ORBITER FAILURE BOARD RELEASES REPORT, NUMEROUS NASA ACTIONS UNDERWAY IN RESPONSE.” (1999): n. pag. Web. 23 Dec. 2014. http://schools.birdville.k12.tx.us/cms/lib2/TX01000797/Centricity/Domain/912/ChemLessons/Lessons/Measurement/measurement1.pdf.

7 “Mars Climate Orbiter Fact Sheet.” Mars Climate Orbiter Fact Sheet. NASA, n.d. Web. 23 Dec. 2014. http://mars.jpl.nasa.gov/msp98/orbiter/fact.html.

8 Hotz, Robert Lee. “Mars Probe Lost Due to Simple Math Error.” Los Angeles Times. Los Angeles Times, 01 Oct. 1999. Web. 23 Dec. 2014. http://articles.latimes.com/1999/oct/01/news/mn-17288.

9 Isbell, Douglas, and Don Savage. “MARS CLIMATE ORBITER FAILURE BOARD RELEASES REPORT, NUMEROUS NASA ACTIONS UNDERWAY IN RESPONSE.” (1999): n. pag. Web. 23 Dec. 2014. http://schools.birdville.k12.tx.us/cms/lib2/TX01000797/Centricity/Domain/912/ChemLessons/Lessons/Measurement/measurement1.pdf.

10 Matlack, Carol. “Airbus: First, Blame the Software.” Bloomberg Business Week. Bloomberg, 04 Oct. 2006. Web. 23 Dec. 2014. http://www.businessweek.com/stories/2006-10-04/airbus-first-blame-the-software.

11 “Airbus A380 – Project Failure Lessons Learned.” Airbus A380 – Project Failure Lessons Learned. N.p., n.d. Web. 23 Dec. 2014. http://www.globalprojectstrategy.com/lessons/case.php?id=23.

12 Matlack, Carol. “Airbus: First, Blame the Software.” Bloomberg Business Week. Bloomberg, 04 Oct. 2006. Web. 23 Dec. 2014. http://www.businessweek.com/stories/2006-10-04/airbus-first-blame-the-software.

13 “Learning from Failure: Engineering Disasters.” Learning from Failure: Engineering Disasters. Materials Science and Engineering, Stony Brook University, n.d. Web. 26 Dec. 2014. http://www.matscieng.sunysb.edu/disaster/.

14 Abkowitx, Mark D. “Vanderbilt Magazine.” Vanderbilt Magazine RSS. Vanderbilt University, Fall 2008. Web. 29 Dec. 2014. http://www.vanderbilt.edu/magazines/vanderbilt-magazine/2008/10/lessons-learned-the-hard-way/.

15 Graham, Ron. “The Engineer’s Companion/FAQ on Failures, Part One.” The Engineer’s Companion/FAQ on Failures, Part One. N.p., 14 Nov. 2008. Web. 29 Dec. 2014. http://www.designnotes.com/companion/failure1.html#03.

16 Abkowitx, Mark D. “Vanderbilt Magazine.” Vanderbilt Magazine RSS. Vanderbilt University, Fall 2008. Web. 29 Dec. 2014. http://www.vanderbilt.edu/magazines/vanderbilt-magazine/2008/10/lessons-learned-the-hard-way/.

17 Graham, Ron. “The Engineer’s Companion/FAQ on Failures, Part One.” The Engineer’s Companion/FAQ on Failures, Part One. N.p., 14 Nov. 2008. Web. 29 Dec. 2014. http://www.designnotes.com/companion/failure1.html#03.

18 Gielingh, Wim. “An Assessment of the Current State of Product Data Technologies.” Computer-Aided Design 40.7 (2008): 750-59. Web.

19 Wang, Xi Vincent, and Xun W. Xu. “A Collaborative Product Data Exchange Environment Based on STEP.” International Journal of Computer Integrated Manufacturing (2013): 1-12. Web.

20 Wang, Xi Vincent, and Xun W. Xu. “A Collaborative Product Data Exchange Environment Based on STEP.” International Journal of Computer Integrated Manufacturing (2013): 1-12. Web.

21 Jackson, Chad, and David Prawel. “The 2013 State of 3D Collaboration and Interoperability Report.” Siemens. Siemens, 2013. Web. 24 Dec. 2014. http://m.plm.automation.siemens.com/en_us/Images/Lifecycle-Insights-2013-Collaboration-Interoperability_tcm1224-210162.pdf.

22 Jackson, Chad, and David Prawel. “The 2013 State of 3D Collaboration and Interoperability Report.” Siemens. Siemens, 2013. Web. 24 Dec. 2014. http://m.plm.automation.siemens.com/en_us/Images/Lifecycle-Insights-2013-Collaboration-Interoperability_tcm1224-210162.pdf.

23 Wang, Xi Vincent, and Xun W. Xu. “A Collaborative Product Data Exchange Environment Based on STEP.” International Journal of Computer Integrated Manufacturing (2013): 1-12. Web.

24 Jackson, Chad, and David Prawel. “The 2013 State of 3D Collaboration and Interoperability Report.” Siemens. Siemens, 2013. Web. 24 Dec. 2014. http://m.plm.automation.siemens.com/en_us/Images/Lifecycle-Insights-2013-Collaboration-Interoperability_tcm1224-210162.pdf.

25 Jackson, Chad, and David Prawel. “The 2013 State of 3D Collaboration and Interoperability Report.” Siemens. Siemens, 2013. Web. 24 Dec. 2014. http://m.plm.automation.siemens.com/en_us/Images/Lifecycle-Insights-2013-Collaboration-Interoperability_tcm1224-210162.pdf.

26 Hardwick, Martin, David L. Spooner, Tom Rando, and K. C. Morris. “Sharing Manufacturing Information in Virtual Enterprises.” Communications of the ACM 39.2 (1996): 46-54. Communications of the ACM. Communications of the ACM, Feb. 1996. Web. 24 Dec. 2014. http://diyhpl.us/~bryan/irc/step/alvarestech.com/temp/0steptools/p46-hardwick-empresavirtual-stepnc.pdf.

27 “Mars Climate Orbiter Fact Sheet.” Mars Climate Orbiter Fact Sheet. NASA, n.d. Web. 23 Dec. 2014. http://mars.jpl.nasa.gov/msp98/orbiter/fact.html.

28 Matlack, Carol. “Airbus: First, Blame the Software.” Bloomberg Business Week. Bloomberg, 04 Oct. 2006. Web. 23 Dec. 2014. http://www.businessweek.com/stories/2006-10-04/airbus-first-blame-the-software.

29 Gielingh, Wim. “An Assessment of the Current State of Product Data Technologies.” Computer-Aided Design 40.7 (2008): 750-59. Web.

30 Neild, Barry. “French Trains ‘too Wide’ to Fit in Some Stations.” CNN. Cable News Network, 21 May 2014. Web. 23 Dec. 2014. http://www.cnn.com/2014/05/21/travel/france-trains-sncf/index.html.

31 “Six Tiny Scientific Mistakes That Created Huge Disasters – World Science Festival.” World Science Festival. N.p., 04 Nov. 2014. Web. 23 Dec. 2014. http://www.worldsciencefestival.com/2014/11/six-tiny-scientific-mistakes-created-huge-disasters/.

32 “2D and 3D CAD Trends in Product Design.” PTC Study (2001): n. pag. July 2011. Web. 24 Dec. 2014. http://support.ptc.com/WCMS/files/136542/en/2d-3d-cad-trends-infographic.pdf.

33 “2D and 3D CAD Trends in Product Design.” PTC Study (2001): n. pag. July 2011. Web. 24 Dec. 2014. http://support.ptc.com/WCMS/files/136542/en/2d-3d-cad-trends-infographic.pdf.

34 Isbell, Douglas, and Don Savage. “MARS CLIMATE ORBITER FAILURE BOARD RELEASES REPORT, NUMEROUS NASA ACTIONS UNDERWAY IN RESPONSE.” (1999): n. pag. Web. 23 Dec. 2014.

35 Hardwick, Martin, David L. Spooner, Tom Rando, and K. C. Morris. “Sharing Manufacturing Information in Virtual Enterprises.” Communications of the ACM 39.2 (1996): 46-54. Communications of the ACM. Communications of the ACM, Feb. 1996. Web. 24 Dec. 2014. http://diyhpl.us/~bryan/irc/step/alvarestech.com/temp/0steptools/p46-hardwick-empresavirtual-stepnc.pdf.

36 Gielingh, Wim. “An Assessment of the Current State of Product Data Technologies.” Computer-Aided Design 40.7 (2008): 750-59. Web.

37 “Mars Climate Orbiter Fact Sheet.” Mars Climate Orbiter Fact Sheet. NASA, n.d. Web. 23 Dec. 2014. http://mars.jpl.nasa.gov/msp98/orbiter/fact.html.

38 Jackson, Chad, and David Prawel. “The 2013 State of 3D Collaboration and Interoperability Report.” Siemens. Siemens, 2013. Web. 24 Dec. 2014. http://m.plm.automation.siemens.com/en_us/Images/Lifecycle-Insights-2013-Collaboration-Interoperability_tcm1224-210162.pdf.

39 Brunnermeier, Smita B., and Sheila A. Martin. “Interoperability Cost Analysis of the U.S. Automotive Supply Chain.” (1999): n. pag. Research Triangle Institute, Mar. 1999. Web. 24 Dec. 2014. https://rti.org/pubs/US_Automotive.pdf.

40 Culler, David E., and William Burd. “A Framework for Extending Computer Aided Process Planning to Include Business Activities and Computer Aided Design and Manufacturing (CAD/CAM) Data Retrieval.” Robotics and Computer-Integrated Manufacturing 23.3 (2007): 339-50. Web. 24 Dec. 2014.

41 Hardwick, Martin, David L. Spooner, Tom Rando, and K. C. Morris. “Sharing Manufacturing Information in Virtual Enterprises.” Communications of the ACM 39.2 (1996): 46-54. Communications of the ACM. Communications of the ACM, Feb. 1996. Web. 24 Dec. 2014. http://diyhpl.us/~bryan/irc/step/alvarestech.com/temp/0steptools/p46-hardwick-empresavirtual-stepnc.pdf.

42 Brunnermeier, Smita B., and Sheila A. Martin. “Interoperability Cost Analysis of the U.S. Automotive Supply Chain.” (1999): n. pag. Research Triangle Institute, Mar. 1999. Web. 24 Dec. 2014. https://rti.org/pubs/US_Automotive.pdf.

43 Isbell, Douglas, and Don Savage. “MARS CLIMATE ORBITER FAILURE BOARD RELEASES REPORT, NUMEROUS NASA ACTIONS UNDERWAY IN RESPONSE.” (1999): n. pag. Web. 23 Dec. 2014. http://schools.birdville.k12.tx.us/cms/lib2/TX01000797/Centricity/Domain/912/ChemLessons/Lessons/Measurement/measurement1.pdf.