Collaborative engineering is not new, but in the fast evolving semiconductor world, it is now the only way to stay alive. Louise Joselyn gathers lessons from guru Patrick Romich
You don’t have to be an engineering management genius to understand that a major key to successful collaboration is effective data sharing, and that this makes data management and data security very important. But data management guru Patrick Romich believes that semiconductor companies, and European firms in particular, have been particularly astute in recognising the need for core systems and processes – and that there are lessons here for others wanting to implement collaborative engineering.
“Semiconductor companies may have taken longer to feel the pain of the current downturn, but in Europe firms seem to have been particularly efficient in recognising the significance of effective collaborative working,” he says. Romich has 20 years experience in managing global software and information operations with Information Handling Systems (IHS), creating CAPS Expert and IQXPERT applications for component and supplier management (CSM) and product content management in the electronics industry. Today he is chief executive officer of Synchronicity, a US-based supplier of data management and IT infrastructure tools aimed at system-on-chip (SoC) design and specifically supporting its requirements for design reuse, complex semiconductor ‘IP’ management and collaborative engineering (see box panel opposite).
Total inter-dependence
SoC design depends utterly on intensive seamlessly collaborative processes, and the semiconductor sector is in the vanguard of those breaking down traditional barriers between people, departments and companies using web-based systems. And what’s important here is that some of its solutions are bound to prove applicable, or at the very least informative, to a much broader industry base. They seem to have generated the tools, the methodologies and the mindset to surmount the problems.
Synchronicity’s customers, for example, include 15 out of the world’s 18 top semiconductor companies, plus a host of ‘IP’ suppliers, electronics OEMs, tool vendors and other organisations, large and small, involved throughout the SoC supply chain. Increasingly, its tools are being specified to provide the underlying infrastructure for collaborative projects, whether inter- or intra-company. Firms such as ST Microelectronics and Philips Semiconductor, for example, are using them to improve design efficiency and designer productivity between engineering groups working in different divisions and different countries, as well as with partners.
Although the technologies behind some PDM (product data management) and CSM tools might be able to handle some of this, Romich is convinced that none can do the job properly. “Their major shortfall is their inability to handle the many iterative changes typically occuring during the SoC design process,” he says. His tools are web-enabled and scalable and can handle a mix of data types so, for example, they can be used hierarchically on large, fast-moving projects involving multiple teams using different tools, to develop complex product families concurrently. They also facilitate data sharing across company, country and time lines.
“Two years ago there was nothing else that would allow corporations to share data world-wide so effectively and securely,” he insists. And it’s important to note that collaborative engineering projects set up using his tools do not exclude small companies. Indeed, the vibrancy of the semiconductor IP sector is due in large measure to the considerable activity of smaller and start-up firms. Supporting software tools can be supplied on a time-licence basis – or ‘seats’ can be free-issued by one of the lead partners.
Beyond electronics
Romich believes that outside the electronics sector, industries using complex electronics in critical components could also benefit from the experience of semiconductor companies. He suggests the automotive industry especially has huge potential which could be realised if it collaborated with the semiconductor sector. “Farming out a specification for a device is not true collaboration. The chip designers need to understand better what’s required, to see a bigger picture,” he says. “Automotive engineers need to plug into semiconductor IP developments to truly understand their potential for reducing duplication and redundancy.”
But there’s more. Data management tools like his are overcoming some of the more commonplace hurdles for any collaborative engineering effort, critically including security and confidentiality. Tools that track who has (or has had) access to what data, that can restrict access to data (internally and externally), that can control and/or monitor changes made to data, that can encrypt or watermark data to prevent copying or theft, are proving invaluable in the high tech sector.
Nevertheless, there remain difficulties – pertaining to business culture. While Synchronicity’s tools have been in strong demand, with semiconductor engineering departments seeing “instinctively” their power (“[They’ve] had to deal with the problems of not having them”), many of the benefits are less easily tangible. Return on investment from, for example, better quality designs, more effective design reuse, avoiding duplicated effort, and even making processes such as ISO qualification easier to complete are, typically, not so simple to quantify.
Tactical or strategic?
Then again, Romich is concerned that buying habits are changing, not just for his tools, but for all IT. “Some firms are no longer buying for the future. Most used to look at least two years hence [and] systems like ours would be a strategic investment, an insurance policy for collaborative design. Now we see just-in-time purchasing.” And he’s not impressed. It’s one thing to refer buying decisions beyond the vice president of engineering up to the board; it’s another to defer purchasing decisions to the point where the IT infrastructure and processes cannot support ongoing viable business.
“Many are not building for growth, but buying in only what they absolutely need immediately,” he observes. If we’re looking for the crucial improvements in competitiveness that come from collaboration, we’d better start thinking and acting a little longer term and understanding the value and the parts we can play.