Silos changing: Design in flexibility

Posted on 18 Apr 2012 by The Manufacturer

This week, Mike Evans, Cambashi research director, discusses new ways manufacturers can design flexibility into their products.

Experts at industry analyst firm Cambashi contribute a regular blog to TM titled Silos Changing. This explores how new software applications enable manufacturers to implement business initiatives in the new economy.

Mike Evans, Research Director, Cambashi

In the 1980’s, designers looked for support for designing flexibility into products used to generate families of parts in CAD parametric modelling – technology typified by PTC’s Pro/Engineer.  While these parametric geometry concepts are still part of the picture, design flexibility today is more likely to mean designing a product that can be enhanced in the field by downloading a new version of embedded software, providing enhanced functionality in the product.

For some time it has been regarded as good design practice for companies to separate research from development. In practical terms this means that research into a material, component or process is a continuous activity. This research may feed into several product families in the company’s portfolio. When the time comes to begin design of a new instance of the product, that’s when the latest proven version of the research results will be used in development.

This good practice has come under pressure however, because product portfolios change more frequently; product architecture is more dynamic; and each development cycle should reconsider and potentially change the implementation technology used for key artefacts in the architecture – for example, functions can be moved between embedded software, silicon and electro-mechanical technologies.

Requirements Management Systems and Systems Engineering and Modelling Tools help maintain the distinction between research and development under these pressures.  The design development team can use these tools to maintain an abstraction of the physical implementation of the design.

Research will deliver better materials, components and processes. Improvements can be quantified to create targets impacting products. For example, the material may be getting 3% lighter for equivalent strength or the manufacturing process 5% faster with a capital investment.  These metrics can be used to decide which research activities to continue and which to either drop, or search for external partners doing better.

The product manager for the portfolio can use these research plans to define a series of instances of products in the family.  Each member of the family will provide different tradeoffs between functionality against wish-list; price; and time to market.  As time passes and research delivers, the portfolio improves generation by generation.

For this to work, the team developing an instance of the product need to know the bigger picture because they are the people who will make a series of detailed decisions that might constrain later instances.  They have to design in flexibility and provide a clear audit trail of their design decisions.

Fortunately, modern design management systems do provide functionality to support more flexible design. Dassault Systèmes’ Enovia; PTC’s Windchill; SAP PLM; and Siemens PLM’s Teamcenter all have the ability to manage the relation between requirements, design artefacts and physical design representations.  Autodesk’s inclusion of a Program and Project Management App in its recent PLM 360 software tools introduction shows that it too recognises the need to address this issue.

Further support is needed from these and other more specialist tools so that more flexibility can be designed into the product portfolio.  A particular problem is how to cope with a product architecture change.  Sometimes, a product improvement removes the need for some components, for example a physical fastener joining sheet metal can be replaced by a tongue pushed through a slot and bent over.

However, the biggest challenge is how to cope with a function implemented with a different technology.  For example, physical knobs and switches can be replaced by a touch sensitive screen and embedded software.  Rainer Anderl’s team at TU Darmstadt have been working in this area and proposed an addition to the generally accepted V model for systems engineering data management.  The proposed W model introduces a cross-disciplinary integration step.

As yet, we are not aware of commercial systems that incorporate functionality to support this concept.  In future blogs we will go on to write about the response to these ideas from the major tools providers.