The task seemed simple enough: “Go get an new toothbrush formy daughter”. The pharmacy had a couple of rows of the traditional ones, butbeing an engineer my eye was immediately drawn to the myriad electric modelsproudly on display. These marvels of technology seem to have come to dominatethe dental hygiene shelves these days, and the choice was impressive.
There were models with ‘intelligent’ charging circuitry tokeep the batteries in tip top condition, ones with ‘programmable speedprofiles’, and models with LCD displays that could display your entire toothbrushing history for the past month, downloadable via mini USB to your computerfor, I presume, scientific analysis by dental professionals (or more likelyparental scrutiny). There was even one model with a UV laser diode embedded inthe head to “sterilize your mouth as you brush”.
Putting the usefulness or otherwise of these oral‘innovations’ aside, the thing that really drives home is how pervasive electronicsnow is in our lives and the products we interact with. From toothbrushes totoasters and toys to tractors, electronic subsystems now play a dominant rolein defining the function and value of the products we buy.
While electronics has been gradually encroaching on newareas of our lives, the last decade, coincident with the prevalence of theinternet, has seen an exponential growth in the complexity of the electroniccircuitry that we come into contact with every day.
One of the consequences of this rapid change is that manycompany production, stock, asset tracking and process control managementsystems are having trouble keeping up with the increasing prevalence ofelectronics systems in the products they produce. Indeed some companies that adecade ago would have had no involvement in electronics, are now findingelectronics production dominating their worlds. And this is putting enormousstress on product management system that were never designed to cope with thecomplexity that electronics introduces.
For instance, many companies have made significantinvestments in large-scale Product Lifecycle Management (PLM) solutions,seeking to create a single view of the complete design process (mechanical hardware,electronics hardware, and services) of a product.
However, most PLM software systems have been developed byvendors with roots in mechanical CAD and 3D modelling software, and targeted atthe physical composition of products. The challenge these solutions face is the”very small scale” and greater complexity of electronic designscompared with the rest of the product into which these electronics areembedded.
The systems are simply not designed to efficiently handlethe number of components and parameters typical of an electronic system, not tomention the programmable elements embedded within the physical hardware.
Not surprisingly, the general approach used is to modelelectronic systems as “black-box” entities within the bigger product.This approach can work when the electronics makes up a small and relativelysimple part of the product.
But as electronics begins to dominate the product makeup,organizations struggle to manage the hundreds or thousands of parts that gointo every electronics design, and increasingly the thousands or more lines ofcode that make these products intelligent, using systems designed to operate ata much more “macro” level.
The situation in product development is analogous to the battlebetween Einstein’s relativity and quantum mechanics. One works well at macrolevel and the other describes things nicely at the micro level. But to reallyunderstand the entire universe, we need a unified theory that works across bothworlds.
In product development our management systems cope with thecomponents and processes that go into the physical design of the product, butstall at the “atomic” level of management necessary to create the electronicsub-systems it contains.
The approach often ends up being one of either “lockingdown” numerous aspects in the electronic design process to minimize riskand allow it to be managed at a macro level, or simply treating the electronicsas black box and managing its development separate from the rest of theproduct. The first approach comes at the cost of stifling (or even killing)innovation in the area of the product that is emerging as the key domain forcreating future differentiation.
The latter creates a disconnect between electronics designand the rest of the product development process that leads to longer thannecessary design cycles and compromised product design.
Neither approach is viable in the longer term. Market forcesare pushing towards ever shorter product life-cycles while customers demand newfeatures and ever more ‘intelligent’ and connected products.
To meet these demands companies must rapidly come up withnew ideas and designs, and streamline the product development process to bringthese ideas to market in the shortest time frame. The big problem is theconflicting nature of the objectives – minimizing risk and production problemsby carefully controlling components and design changes, versus encouragingdesign innovation to drive future product differentiation.
The first step in solving a problem is knowing where the problem lies and who owns it. In this case the responsibility for finding a solution is largely in the electronics design domain, where the detailed and fluid nature of electronics design data is often incompatible with broader production data needs.
For example, when we’re designing and manufacturing our electronic toothbrush (the one with the mass spectrometer in the tip to measure the acidity of our saliva and warn us that we’re eating too much sugar), the only thing we really need to know about the electronics at the product level is is that the correct version of the electronics sub-assembly has been manufactured and will be delivered to the factory in time to be loaded it into the casing.
The task of getting the electronics sub-assembly designed and manufactured, however, requires detailed information about all the many components that make up the boards, including many intangible items such as software and firmware loaded into the system after manufacture. The key to streamlining the overall product design and production process is to have single and coherent point of management for all the electronic data that then feeds only the essential information into the product production process at the macro level.
The fly in the ointment here is that our current electronics design processes are severely fragmented. Typically we use completely different tools for different parts of the design, each storing it’s part of the design data in its own database and generally communicating in only the most rudimentary terms with the other tools and databases in the design chain.
The fragmented nature of the design tool chain makes it very difficult to put together a single coherent picture of the overall electronics design status necessary to feed wider enterprise systems.
The first step in reconciling electronics design with the wider product development task is to take a step back from our current fragmented design processes and the subsequent grab-bag of tools required, and to seek to unify what are traditionally treated as disparate design processes (Figure 1, below ).
We need a new generation of design tools and processes that recognise at the outset that an electronic product not just a PCB, nor a set of programmable gates, nor software running on processor, but a combination of all these things and more.
Tools are emerging that recognize the importance of design unification and the need for a single, coherent data model to underpin the entire electronics design effort. This is important, not only because it makes the process of design easier, but because it provides a platform on which to build the data management capabilities that will be necessary to build the next generations of smart, connected electronic devices.
Once we adopt a unified approach to electronics design, many of the issues that plague electronics development today, such as change management and versioning, the creation and management of hierarchical projects, and release management can be addressed at a design data level within the design tool and without the design restrictions usually associated with imposing an external data management system that needs to address and synchronise data from multiple design domains.
For example, it’s common during a design cycle to need to modify a component at the board layout stage for packaging or availability reasons. But unless strict change management procedures are in place, it’s easy for the base schematics to get out of synchronisation with the board design. If the BoM for manufacture is generated from the schematics, then the wrong part can be ordered leading to problems and delays during manufacture, and subsequent flow on effects for the entire production schedule.
With a single, unified model of the design (Figure 2, below ), and indeed the components that make up the design, changes in one realm – the board – are automatically propagated to other realms – the schematic source. Basically, it’s all the same data set. The system can then ensure that any data generated for a specified purpose, such as a release to manufacture, is valid across all domains and completely synchronized and up to date.
At the macro product production level, this has the advantage that any data extracted for the purpose of determining the lifecycle status of the electronics is always current and correct.
Figure 2. A platform approach to design, where a unified data model underpins the entire design, makes it easier to manage design data across multiple design domains, and provides the scalability to extend design into new areas as needed.
It’s a pretty safe bet that the extent and value of electronics will increase in the next generation of devices we create. Electronics will move into new areas and grab more of the mind share of companies in all sectors of product design, not just those we think of as ‘electronics’ companies today.
It’s imperative that electronics design be completely integrated with a company’s wider product development processes so that design and manufacturing cycles and can meet the needs of a demanding market. But it’s impractical to attempt to bring the complexities of electronics design and manufacture directly into management systems targeted at the higher level task of controlling and tracking the wider product lifecycle. Instead, robust data management must be made an integral part of the electronics design process itself and the design tools used to carry out the process.
For any system to effectively manage design data, it needs to have full access and control over that data. When multiple, disparate tools and multiple sets of data are involved, creating an effective system for managing the overall data flow is next to impossible without severely limiting the designers freedom to make design changes.
So a prerequisite to building tomorrow’s electric toothbrush, the one with a full diagnostics suite in the head that uploads data directly to a cloud-based medical system, is to move away from a tool chain approach to electronics design and towards a platform-based approach that unifies all design data within a single, coherent model.
The next generation of intelligent, connected electronic devices will not be defined purely by the hardware in the box, nor by the firmware and software programmed into it, but by the complete user experience created by the device and the ecosytem it connects to.
Increasingly electronics hardware, programmable hardware and software work as one to dominate our experience of a product. Our design tools need to reflect this unity.
Rob Irwin, Product Manager, Altium Limited, has a Bachelor of Engineering (Electrical) from the University of Sydney, Australia. He has over 20 years’ experience in the electronic design industry including several years as editor of Australian Electronics Engineering.