Managing your products from cradle to grave

It is generally accepted that product lifecycle management (PLM) describes a series of business processes, enabled by application software, proven to generate business value in a variety of industries when tracking a product from ‘cradle to grave’. The definition is a little vague, not least because even the consultants are finding precision difficult. At a recent seminar run by analyst Cambashi, findings from a small user survey broadly showed that few users know what PLM even stands for, and even those that did were hard pressed to come up with a definition. And it’s not so different on the vendor side. At the recent COFES (Congress on the Future of Engineering Software) in the USA, the group given the task of defining PLM found it far from easy, and the result looked more like a vision statement of requirements than a system specification. Fact is though that PLM is now centre stage in CAD/CAM, having ousted ‘collaborative engineering’ as the current marketing buzz. As was the case when CPC (collaborative product commerce) bested PDM, the software has changed little, and it is likely to remain that way for at least another six months, or until the next major vendor release cycle comes along. But the coining of a new acronym is a bit like a self-fulfilling prophesy – sooner or later the big CAD/CAM providers will offer tight integration with the range of applications implicated. With several companies from various sides of the manufacturing IT divide vying for supremacy, boundaries are becoming blurred and some software categories are in danger of losing their separate identities. The most intriguing suggestion comes from Bernard Charlès, head of Catia developer Dassault Systemés. He advocates that, in discrete manufacturing, “SCM will vanish into PLM and ERP.” And if you search the web for ‘web-based PLM’ you turn up all the usual manufacturing application suspects – SAP, Mapics etc on the ERP side, with EDS, E-vis and so on beginning to creep in. As Peter Thorne, senior consultant at Cambashi, says: “While CAD and PDM are hugely important in achieving the PLM vision, it is easy to overlook the enterprise applications side of the business.” The transition from ‘as designed’ to ‘as built’ typically involves production planning and control systems (from ERP to MES). These are not only the source of material and batch traceability information, but also may provide production people with discretion to allow component substitutions. The use of a ‘better’ component rather than delaying production, is a valid manufacturing strategy. Serious horses for courses Then there’s the transition from ‘as built’ to ‘as installed’ and ‘as maintained’. Covering these areas there are maintenance management and asset management systems, which often have their own source of product information, and schemes for using it and updating it. Thorne: “I have no problem if the gas engineer servicing my domestic boiler refers to original design information (I expect he carries the significant engineering changes in his head). But I’d prefer the engineer servicing an aircraft I’m about to fly on to be using up-to-date ‘as maintained’ information.” And hence PLM. In the process industries you can optimise the supply network without some product information. However, in aerospace, a bolt fastening the engines to the wing has geometry, tolerances, materials strength and a host of other properties that must be taken into account before anyone in the supply chain can substitute an alternative bolt. For an SME whose only interest is ‘as designed’ it is difficult to reconcile the all-embracing nature of PLM with the business processes of SME suppliers of whatever kind further down the supply chain. Fortunately, the major CAD/CAM vendors offering PLM solutions so far provide a modular approach so ‘you pays your money and you takes your choice’. A company that can just about be called an SME is Alstom Train Services (ATSL), which grew out of Alstom Transport in the mid-1990s. Following the privatisation of British Rail, a significant business opportunity emerged in the provision of ‘asset management’ and lifetime product maintenance in after-sales. Alstom Transport in the UK, which designs and builds trains, has around 350 active users of EDS’ Teamcenter PDM software suite (previously Metaphase) used to manage all data associated with new product design. ATSL, whose numbers are more modest, was able to look at what the parent company had achieved with its PDM, as applied to the management of product data associated with repair and maintenance. It became clear that the requirements of a service environment differed from new build in some significant ways. First, the period over which data was actively required by product development was typically three to five years. However, ATSL would need to keep data available for up to 30 years to provide maintenance throughout a train’s working life. Second, the new-build installation of Teamcenter involved many users in few sites, whereas the ATSL services users would be far fewer in number and spread over many more and smaller sites. Third, with ATSL, source data would be held at multiple locations rather than at a single site, requiring a federated approach to database management. Its project requirement was to understand the product data flow from cradle to grave and to reconcile the service activity with this data flow. The resulting process schema envisaged the product data flow in terms of a fully integrated single business process, which included the new build data, the service data and some Alstom central data sources and repositories. This produced a federated PDM environment within the UK and France, as well as a measure of interfacing with the ERP system to manage the actual train configurations. It’s a big example, but it makes the point: the system has brought significant efficiency improvements. Component identification is now much faster because all relevant documents and drawings can be viewed online. In the past, referral back to the original engineering department was often required, with the result that response times could be as long as several weeks. “The system implementation has stayed below budget and on time for all major milestones,” comments David Millin, ATSL PDM project manager. “This is against a background of other IT infrastructure projects which have run significantly over on both of these key project measures.” The system is also serving as an enabler for future developments – in particular, e-maintenance via wireless and on-train diagnostics. As the data is made available electronically (ePLM), it can be integrated with other technologies to provide a comprehensive maintenance environment. For example, it gives ATSL the chance to be proactive as, using fault-recording systems, it begins to monitor trends across the entire fleet. By monitoring the train’s vital functions, if they start to wander off their parameters, people know about it before the train comes in. You can see from this that real PLM cannot be an entirely ‘out-of-the-box’ activity, and that it requires significant thought and effort – business process and IT – over and above the implementation of PDM. Although introducing PDM doesn’t change the maintenance regime, it does start to enable other technologies which do change the way maintenance is done. The major impediment to widespread adoption of ePLM will be the expense and the potential security issues. As John Maynard, mechanical engineering manager at Thales Defence, says: “Eventually, something like ePLM will work very well, but the amount of effort that has to be applied to it is considerable. And as with most things like this you don’t really know the effect it’s going to have on the whole culture of your business.” His advice: “It’s easier to introduce these things in a fairly small way. It’s also easier to use neutral ground (ftp sites or third party hosted websites). As soon as you open your systems to the outside world your security suffers and potentially the integrity of your data is at risk.” Time to take more interest.