This article discusses the shift to product lifecycle management (PLM) systems by various mechanical engineering companies. Systems engineers and information-handling experts are joining forces to get a grip on the information explosion, thanks primarily to the timely convergence of systems engineering with digital design and development. PLM supports the extended enterprise. The rationale behind using PLM is to ensure that the ideas and information driving the development of today’s products incorporate best practices and everything learned right up to the product-release date. The rapid increase in electronic controls and software that are being built into key auto components requires that mechanical engineers and electrical engineers work ever more closely together. This highlights the need to integrate the very different approaches to development that the two disciplines use. One of the key functions of PLM is to make sure all the data in those analyses are retained, not just the conclusions.
Systems engineers and information-handling experts are joining forces to get a grip on the information explosion, thanks primarily to the timely convergence of systems engineering with digital design and development. And aptly so, since for two decades new product design has been a major pain point in information handling.
In the world of mechanical engineering, one of the biggest sources of the data explosion is new product development. Over the past two decades, the information fl ow has been transformed from a trickle of engineering drawings and scattered test data, all of it on paper, then to a few dozen 2-D CAD fi les, and today to a digital tsunami that touches every part of the organization.
These digital tools were originally intended for what might be called housekeeping in new product development—gathering and organizing engineering data, and simplifying its retrieval. As the power of the tools was grasped, they were set to work doing things that had previously been impossible, or at least not cost-effective. These tasks included digital prototyping, cataloging legacy data, tracking customer-account information, storing know-how, and much more.
Information sources span the enterprise from concept development, through simulation and analysis, prototyping and finally to compliance with end-of-life disposal regulations. Users include purchasing, the enterprise resource planning system, finance, marketing, manufacturing engineering (ergonomics, quality assurance, and productivity), and field service, plus customers, suppliers, business partners, and distributors among many others.
These gushers of information reveal previously hidden small but profitable design opportunities, detect flaws earlier in the development process, recognize dead-ends sooner, winnow out many prototypes, and ultimately smooth out and accelerate manufacturing development. The downside of all this is equally clear: too much of a good thing.
PLM supports the extended enterprise (customers, designers, supply partners, etc.) from concept to the end of life of a product.
The answer has been a powerful new shove for product lifecycle management, or PLM.
Every successful business strategy needs the coherence of a sound definition. CIMdata defines PLM as a strategic business approach that applies a consistent set of business solutions that support the collaborative creation, management, dissemination, and use of product definition information. PLM supports the extended enterprise (customers, designers, supply partners, etc.) from concept to the end of life of a product.
Beyond discrete manufacturing, where PLM started, it applies equally well in the process industries and in architectural-engineering-construction. In the process industries, PLM is focused on the plant itself (such as a refinery or a power-generating station); in AEC, it is focused on a building.
The rationale, PLM project managers say, is to ensure that the ideas and information driving the development of today's products incorporate best practices and everything learned right up to the product-release date. Before a company can leverage its information, it must keep track of it. By integrating people, processes, business systems, and information, PLM can be the answer to that challenge.
A company's new product digital data starts with conceptualizing. As the product idea takes form and enters development, the cascade swells with specifications, CAD models, results of tests and analyses, bills of materials, orders for tooling, and so on.
This information is reused, reformatted, and replicated in dozens of databases and decision points in purchasing, finance, marketing, manufacturing engineering, and field service. It's also used to populate a manufacturer's enterprise resource planning system and is depended upon by customers, suppliers, business partners, and distributors among others.
With all of these demands, sound management of information, and PLM in particular, pay off by finding needed information and avoiding its recreation whether as new CAD drawings or data re-entry.
A big part of the information explosion stems from striving to come up with compelling new products amid global competition, which has driven the rapid expansion in the use of simulation and analysis software in the engineering industry.
Other big drivers are health, safety, and emissions regulations, and the fear of litigation.
“All the data will never exist in a single location,” said Christopher Hoffman, a systems engineering process leader at Cummins Inc., the diesel engine manufacturer in Columbus, Ind.
The PLM challenge at Cummins, he said, is that good product-development processes are available. “But individual engineers and technical people at the everyday working level frequently face fragmented and uncoordinated views of data and process support that he or she needs,” he said.
“The individual too often must manually re-enter data for different activities, and can only hope that the data properly aligns with data that others are using,” Hoffman said. “Such a work method is prone to process and data inaccuracies. Traceability is poor, and process efficiency suffers. It is a real challenge to effectively integrate process, data, work templates, and program management in a practical fashion.”
Systems engineering tools at Cummins provide accessible, convenient, and configurable work environments that appeal to both systems- and non-systems engineers, Hoffman said. The work of these engineers includes managing documents about departmental deliverables and evidence of delivery, requirements for traceability and critical parameters, failure mode effects analysis and risk management, systems validation and verification, and Six Sigma quality assurance.
All of these are components of PLM and it is the confluence of new product design and systems engineering that is driving the adoption of PLM. As with any new technology, good tools in the users’ hands support and eventually compel adoption.
In the Know
Dealing with tribal knowledge has been a significant issue for Bissell Homecare Inc., a 135-year-old floor-care appliances company in Grand Rapids, Mich. Tim Field, manager of mechanical design and CAD, and Alan Krebs, lead engineer for global technology and innovation, explained how Bissell uses knowledge-based engineering to extend the company's tribal knowledge to its global business. This was triggered by Bissell's rapid expansion overseas in recent years. Bissell has manufacturing operations in China, Korea, and Mexico, as well as the United States.
In manufacturing, tribal knowledge is unwritten but valuable information that accumulates and is shared within a work group—know-how—but it is not often shared with others, at least not freely. The PLM challenge in dealing with tribal knowledge is that it lacks verification by analyses or other data, and is poorly linked to the enterprise's information flows. Knowledge-based engineering, or KBE, ferrets out tribal knowledge with a combination of CAD, object-oriented programming, and artificial intelligence.
Krebs said that knowledge-based engineering “captures our global tribal knowledge with virtual models. This smart (and simple) geometry makes it easy to create ‘what-if’ designs that can be readily tested with simulation and analysis. A spreadsheet is used to drive the CAD geometry making it easy to use for all non-CAD users,” he added. What Bissell engineers get from this is “consistent and speedy creation of mechanical layouts, a push toward modularity, and the implementation of global design standards with tighter control, with more consistent design and engineering procedures.” This is yet another form of a single point of truth.
“Data reuse is also much greater,” Krebs noted, as opposed to recreating or redrawing with its penalties in time, cost, and design consistency. Along the way, Bissell engineers have firmly linked knowledge-based engineering with systems engineering.
“These gains allow performance breakthroughs to be readily shared across the global organization,” Krebs added. His background includes key roles in Bissell's global technology and innovation unit and in new-business development.
Bissell's knowledge-based engineering model for its upright vacuum cleaner includes over 300 direct and indirect performance characteristic values. These values control the specific geometries that contribute to best performance. In the past it would require weeks of effort to specify the desired values and build the 3-D mechanical layouts. With knowledge-based engineering, a 3-D mechanical layout can now be generated in less than a day.
Silos of Expertise
According to Len Wozniak, process and tool systems architect for electronic controls and software at General Motors Co., a particular challenge is the tendency of different parts of a company to operate in silos. It is especially true of mechanical engineering and electrical engineering departments. As he laid it out, the problem has been the lack of a multidisciplinary orientation, tools, and capabilities in the development of the electronic controls in, for example, vehicle steering, braking, speed control, and similar systems.
Wozniak said his team has achieved some documented successes, which have big implications for PLM strategies. Among the challenges GM is overcoming are the lack of a multidisciplinary orientation, huge differences in the ways MEs and EEs report their design progress, and the metrics they use.
Engineering projects are managed with phase gates—points in development requiring a decision to proceed or not. In any project, decision points for MEs and EEs rarely coincide; this can greatly complicate the timing of management decisions.
The rapid increase in electronic controls and software that are being built into key auto components requires that MEs and EEs work ever more closely together. This highlights the need to integrate the very different approaches to development that the two disciplines use.
The internal engineering structures, or silos of expertise, add complications. So does the unfamiliarity of the typical auto industry ME with the ways in which software development is managed. In electronics, that process is product lifecycle engineering (PLE); roughly speaking PLE is PLM's counterpart in electronics.
Automotive product development traditionally focused on mechanical components. The main concerns were fit, function, and durability; until the advent of “mechatronics,” electronics and software were involved only peripherally.
Wozniak said two big areas where GM has had success have been in reducing engineering costs for electronic control units (ECUs) in brakes, steering, etc., and a significant reduction in warranty claims.
Engineering costs per ECU dropped by 26.5 percent the first time PLE approaches were married to customary ME methods. Engineering costs per ECU fell a further 9.75 percent the second time, he said.
The cost of warranty claims for all vehicles sold in the past seven years fell to 0.3 percent of vehicle cost from 1.07 percent.
Challenges remain in both tools and culture. Wozniak said tools are needed to manage parallel streams of development that occur when PLE and PLM are both in use. On the cultural side, he said, “While all product teams welcome the quality and cost benefits of PLE methods, few understand how they work.”
There was a time when many believed that, once everything went digital, everything would be simple for engineering departments. It turned out, of course, that going digital was anything but simple and straightforward. It made the world more complex and richer for it.
So much more could be done with computers and software than anyone had expected—simulation and analysis, for example, to slash the number of prototypes and compress manufacturing tryouts. One function of PLM is to make sure all the data in those analyses is retained, not just the conclusions.
There was a time when many believed that, once everything went digital, everything would be simple for engineering departments.
Instead of fading away, specialties and divisions of expertise multiplied. Looking past the very real technical challenges of data connectivity and interoperability, from the PLM and informationhandling standpoint, silos are a big systems-engineering issue because their organizational charts are dynamic and their workflows ever-changing.
Today, companies are striving to enhance the value of the information they hold, to prevent its loss, and to find innovative ways to use it. The challenge is that critical information originates in many different departments, locations, and formats.
How does a company keep its engineers from redrawing the wheel? How does it take a good practice from a plant in the American Midwest and make it available to branches around the world?
Many companies say they are turning to PLM systems to do it.