Professional Engineering Magazine : Building in the brains

Return to:

Building in the brains

Take off-the-shelf modules, customise them and, hey presto, you’ve got a digital prototype in double-quick time. John Pullin reports from Texas

The difference a few months makes is pretty astounding. Engineers James Carver and Ivan Charamisinau of Animage show a mock-up of what their veterinary diagnostic imaging system, Fidex, looked like in January, and then a view of the machine in March. The first looks like leftover bits from an unsuccessful attempt at self-assembly furniture; the second is a proper machine of the kind you might expect in a vet’s surgery.

But the point they’re making isn’t actually the physical transformation of Fidex, a system that combines three different types of imaging in a single set-up so you don’t have to move zapped moggy from machine to machine to see what’s going on inside her. It’s that, once you’ve got the concept for a product proved, doing the physical configuration isn’t actually the difficult bit.

Carver and Charamisinau were at National Instruments’ annual Texan tech fest, NI Week, this month to explain how they, with limited experience of embedded system design, had managed in just a few months to develop a benchtop prototype to control the X-ray source, the X-ray detector and then the motion system, using NI’s Labview graphical programming language.

And, having integrated complex control and data acquisition elements into a functional prototype, they were able to use the same codes inside NI hardware to make a first preproduction system, using a single board designed for embedded machine development.

The brains of the operation: Data acquisition modules linked to programmable systems are at the heart of NI’s technology

The point the Animage pair were aiming to illustrate was that using off-the-shelf software and hardware modules and then customising them to their own application had meant they didn’t have to develop their systems from scratch. “We estimate that this saved three man-years of development time, or $300,000 in labour costs,” they said.

The point from this that NI is wanting to make, says Arves Stolpe, marketing manager for the CompactRio range of data acquisition modules, is that what can be done for cat scans of cats works also for lots of other complex systems where designers want to build diverse or multiple data sets into a test regime and a control system.

He cites as an example Solix, a Midwest developer of algae-based biofuels, where the big challenge is monitoring vast shallow ponds where the algae is grown to ensure that the balance of water and carbon dioxide is kept at the optimum. Solix has prototyped a system that uses wireless nodes to monitor and control its assets. On the scale of the development planned, claims Stolpe, it’s hard to see what else could do the job.

But there is a wider lesson in all of this for product and system developers, NI reckons. What the new integrated software and hardware tools enable designers to do, says Labview product manager Rick Kuhlman, is to cut to the chase more quickly by getting digital prototypes up and running faster and then moving on from them into functional prototypes.

And he cites some “golden rules” for prototyping. “For a start, you have to recognise that ideas are cheap,” he says, “and that the expense comes when you have to test or verify them to see what has real economic value.” Systems that can cut this cost or the time to do this apply to virtually all firms.

The three-in-one system that enables a cat to be cat scanned

Second, at the digital pr++ototyping stage, you need to put in “just enough” work: not too little, but not too much either. “You don’t want to go too far down a path to show that something works if no one’s going to buy it.” You want to test the feasibility and then open it up to customers to get feedback before committing more cash to it.

It helps too, says Kuhlman, to build modularity into systems: customers will always want extra features or their demands will alter subtly: “Anticipate that there will be some ‘future creep’,” he says.

Animage’s three-in-one diagnostic tool for animals is an example of a fourth dictum: design routines into the prototypes that you’ll be able to reuse in the final product wherever possible, says Kuhlman. It saves time and money.

If several of these ideas seem to be about money, then that is the focus for Kuhlman’s fifth golden rule: don’t focus on cost reduction too early in product development. “You can easily talk yourself out of whole markets if you do your cost optimisation from the outset,” he says.

“By all means optimise once you’ve got a product. But at the prototyping phase you’re aiming to build the brains into the system.”

	Previous focus story	Next focus story
	Back to contents	Email us your views

www.profeng.com The home of Professional Engineering on the web

Professional Engineering

Related links and resources

Building in the brains

www.profeng.com The home of Professional Engineering on the web

Professional Engineering

Search our back issues

Related links and resources

Building in the brains