Early stage design

by Peter Thorne
@Cambashi_Peter
LinkedIn

Early stage design decisions have big impacts. Early stage design studies guide these decisions, for example, decisions about layout, material choice, make-or-buy, initial definition of subsystems, production methods and many others. These choices form the framework for the rest of the project, and so have a direct impact on costs, successes and problems later on. Software tools that support better early stage design decisions offer big ROI potential.

Value and volume

The value is up front, but the volume is downstream. Those early overall architecture decisions have major implications for the rest of the project. Arch, suspension or cantilever bridge? Gas turbine or diesel ship engines? Single integrated unit or three connected devices? Plastic, sheet metal or machined? In contrast, later in the project, there are thousands of detail design decisions, each essential, but, if the early stage decisions were good, these later decisions will probably have just local ramifications.

Big ticket has… “been there, done that”

Development workflows for big ticket projects such as new automobile models, aircraft, ships and petro-chemical plants have, for decades, had specific phases for early stage design work with names like ‘preliminary design’, ‘concept development’, ‘front end engineering design’ or ‘initial design’. The teams involved have specialist software to help guide choice of the key parameters – the length, beam and number of engines for a ship, or, in process design, the pressures and temperatures that will optimize yield for an expected range of feedstock purity. Engineers who recommend diesel or gas turbine engines for a ship; or propose the number of distillation columns in a process plant; or sign off new lightweight materials for a car or plane make the choice based on the results from specialist software plus a healthy dose of experience.

What about everyone else?

Away from big ticket projects, software to guide early stage design has been less visible.  Many teams use detail design tools, plus spreadsheets. Component and material suppliers often invest in order to get involved at this stage – if they can get their products selected at the early design stage they will almost certainly achieve better margins than will be needed to win a sales battle later.  So these companies find ways to make it easy for design engineers to select their products early in the project.

Now, growth of systems engineering and new simulation and analysis capabilities are driving more capability into the hands of early stage developers.

So in this article, I’ll characterize early stage work, and then select a few examples of engineering software tools which are addressing these needs.

m201_prod_34
Figure 1: From concept to volume manufacture

The broad principle is the same across the spectrum, from high-cost one-offs such as buildings, ships, and process plants, all the way to low-cost, high-volume products such as disposable razors and even toothpicks (wood or plastic? Individually wrapped or multipacks? 50mm or 60mm?).  Early decisions, viewed in the context of the project, tend to be big decisions. These decisions establish the framework for future work, and are expensive to change. Yet in most cases, a relatively low share of the total project development resource is allocated to making these decisions.

This is why, when viewed from the stratosphere, product development still tends to look like a one-way process, in which multiple design studies are used to guide choices between alternatives, one of which is chosen for development (fig. 1 shows a volume manufacturing example).

It’s not a one-way street

Move in from the stratosphere, and of course the process is far from one-way. Agile methods enable development engineers across the whole lifecycle to try relevant ideas in every sprint. Also, systems engineering is helping multi-technology product development teams divide tasks across mechanical, electrical, fluids, and software.  This partitioning enables concurrent work on multiple subsystems.

So what is different about early stage work? Of course there is uncertainty.  Also, setting up the context and environment needed for the new design can be time consuming and can feel like overhead. The definition available for all the current design ideas is very limited, so it can be difficult to evaluate those ideas. Some ideas may turn out to be requirements (connected to internet, energy consumption less than x,…), and the early stage designer has to seek out all the implications.

The ability to perform analysis and simulation based on limited inputs used to be confined to big ticket projects, and special cases where someone built a useful spreadsheet.  Now there are tools relevant to every product development team. Some help convert broad brush into more detail.  Some enable more evaluation of broad brush definitions.

Tools, what tools?

There are many types of tool, some apply across a broad range of product types, some are more industry or technology focused.  To give a taste of what is available, I’ll pick 8 tool types, and give an example of each. Support for early stage design decisions is at least part of the purpose, and sometimes the dominant part of the purpose of these tools.

Is this a trend?

Several trends in the marketplace are converging to help better serve early stage design. It would be hard to say that market demand for early stage design tools was the factor driving these trends.  Key among the relevant trends are:

  • Democratization. Software vendors always seek new users. Early stage designers include senior people with a range of responsibilities, so will never be full time users of any system. Easy, discoverable, intuitive UIs make tools accessible to these people.
  • New simulation/analysis. The fact that simulation and analysis originally was something that happened downstream has been a barrier. The combination of easier modeling and integrated analysis allows early stage designers to run simulations earlier in the project, trying and testing ideas, using agile methods to refine requirements and then try new ideas.
  • Growth of systems engineering. 50 years ago, SE happened in most aerospace projects, many military and some automotive projects, and it developed a reputation for being a documentation heavy, bureaucratic approach. Even 20 years ago, tools were largely requirements management, except in software development, where structured analysis, and modeling was well established, and tools could be linked to code generation and test.  But now, systems engineering tools are relevant for multi-technology products across most industries, and can be key to managing complexity.

Where next?

Strong take up of tools in early stage design will make it hard to recruit design engineers into an environment without these tools. These hard-to-recruit design engineers will have become used to projects in which early stage decisions are almost always good decisions. For them, an environment where it’s not unusual to discover big problems late in a project will feel like a backwards career move.

The role of component and material suppliers also looks set to grow. Every potential supplier wants to be present and influence the choices made during early stage design. Online catalogs and sales configurators were the starting point, but now design tools from suppliers, often low cost or even free, offer much more, for example, automatic generation of bid documents. Of course, designers have to accept that the best functions only work when applied to the parts of the design based on products from that supplier.

Another axis of growth is the product lifecycle – from first idea through to recycling.  IOT is driving a downstream rethink – how should design and development engineers be connected to products in the field? But there are also opportunities upstream, driving out cost by helping companies achieve right first time, every time in the big, early stage decisions. Simulation capabilities integrated with systems engineering will provide the capabilities; easy-to-use interfaces will make these capabilities accessible to the right people. ESI Group has been active in the closely related ‘simulation-driven-design’ area, its acquisitions of CyDesign, Amoeba (Presto), Ciespace, PicViz (Inendi), Mineset, Civitec and recently SciLabs all contribute to a growing footprint of early stage design capabilities.  Ansys AIM also targets this market, and its SCADE suite, originally part of the ANSYS acquisition of Esterel, opens fascinating questions about systems engineering and software modeling in simulation.

Your say

What are your thoughts about early stage design? Will it become standard practice? What is your preferred tool? Share your views in the comments below, or contact me directly via e-mail.

Disclosure: Autodesk, Dassault Systèmes, PTC and Siemens are Cambashi clients.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s