Viewpoint: Bringing it home to roost
01 February 2022
The electronics industry has taken its worst battering in living memory over the last few years. While the industry is still scrambling to achieve previous production levels now two years into an ongoing pandemic, is there anything that can be done to alleviate the problem, asks Colin Funnell, Development Engineer at embedded systems expert, Hitex? Surely after all these years, ‘the system’ has evolved to be robust?...
This viewpoint was originally featured in the February 2022 issue of EPDT magazine [read the digital issue]. And sign up to receive your own copy each month.
The basic fact of the matter is that, for years, we’ve all grown accustomed to ‘getting what we want, when we want’. This applies to a whole range of experiences, from online shopping to industrial services – and companies have streamlined themselves to take advantage of this. If anything, we’ve been drifting to a less robust model by relying on so many others. We’ve experienced disruptions in the past, say a silicon fabrication plant goes out of action, which have had effects on end markets for up to a financial quarter. But now, a perfect storm of events has all but crippled the electronics industry. The list of problems goes on and on…
Modern times & modern problems…
• A pandemic has impacted everyone in the manufacturing and supply chain. At the same time, fabrication plants have had fires and suffered from natural disasters.
• COVID-19 has shifted normal consumer patterns. Fewer cars are wanted, but more computing devices are needed to enable working from home. This has resulted in an upheaval in supply chain forecasting, most visibly in the automotive sector, with chip foundries trying to adapt to rapidly changing demands.
• Whether you are in favour of Brexit or not, this too has contributed, as obtaining components internationally has been made more difficult.
• Just as you might have thought that the run of bad luck couldn’t possibly continue, shipping was disrupted by the Panama Canal being wedged shut.
• And you can virtually guarantee that by the month you read this, there will be some other incident to add to the list...
These may be extreme times, but this succession of problems has helped highlight the risks of off-shoring so much of your operations. By having so many links in the production chain, you reduce your control and influence over what can be done. Although the most obvious problems have been with physical components, a similar argument can also be made on outsourcing development work too far afield.
We have seen remote working taken to new heights. With a plethora of computer tools available, it is now possible for people to provide useful work from home across a range of job functions. By extension, the remote aspect doesn’t have to be an existing worker’s home, it could harness a worker from anywhere in the world. The age-old question of employing someone locally versus outsourcing to a cheaper region now has a whole new angle. Is it time to re-examine the need for national, local, hands-on expertise?
Things have come a long way since the first years of outsourcing contract manufacture of PCBs (printed circuit boards) to China. A wide range of manufacturing grades are available with appropriate costs. Quality and capability can be up to world-class standards. With more consistent manufacturing cycles, UK companies are partnering up with Chinese manufacturing plants to improve traceability and end quality.
However, the communication and time zone difference can still be a problem, especially where low-volume development prototypes are involved. The trend is still to produce these on-shore anyhow. Re-shoring can only hope to compete on quality, not price, although the gap is narrower than it used to be – and if the cost of replacing failed boards is very high, it may still make financial sense to manufacture even high-volume systems in the UK. Given the choice, we would manufacture cheap, non-complex LCD (liquid-crystal display) carrier boards in China, but complex hybrid vehicle development platforms in the UK.
Design work is a different story. Here communication and consistency are key, together with quality management systems, such as ISO9001:2015. This is why designing PCBs is a prime candidate for domestic work: local project groups, working together geographically, within a quality system, will always result in a higher quality result than disparate groups, working independently, via often inadequately defined specifications.
As seen so many times, requirements can be the let-down of collaboration – locally or abroad. With the push for low-cost outsourcing, and the rise of communications tools over the years, requirements specification and writing skills are starting to slip away rather than reinforcing clarity. Instead, the continual messaging approach driven by mobile telephone culture seems to be the way to steer progress. With multiple chat platforms, and the lack of formal recording, traceability has fallen by the wayside.
Components have their geo-political issues too. When key project components, electronic or otherwise, have international origins, the more localised the design and the technical support is, the better the design will generally be. Modern pseudo-support will redirect you to a forum or bot which may not answer your question fully – but may provide just enough to keep design work progressing. Local FAEs (field application engineers) are worth their weight in gold, especially those who can offer samples to help get your prototypes up and running.
The collaborative aspect needed for software development can be even more pronounced. The specification and communication must be very tightly defined, together with finding a common cultural approach to development. For more straightforward work, outsourcing abroad can bring considerable financial benefits, but the trickier, more involved bits of development benefit from closer working in groups as a team. Again, we have got this down to a tee at Hitex, with all complex development being carried out with neighbouring teams in-house or with partners in the UK.
The past 18-months (and counting) have been horrendous for any form of electronic manufacturing – and this seems set to continue into the near future. A PCB parts list (BoM or bill of materials) can almost be seen as a massive risk analysis, with each component type now having a probability of availability, resulting in an overall uncertainty that the product can even be built.
Many companies over the years outsourced manufacturing so they can focus on their core business for an incredibly lean just-in-time approach. But by not even keeping stock of key components, this translates business cost savings to business risk – a gamble which many are paying for now.
A trade-off might be to modularise a design. A CPU module contains a number market sensitive devices – mainly the CPU itself and memories. Handling your own stock of high technology components may not be palatable, but holding stock of modules might be. Module manufacturers will have greater purchasing power for their volumes, but may still not be invulnerable to the component markets. It might not be a magic bullet, but it could still help reduce your build management overhead.
Design for availability
Due to the global complications we’ve had, circuit design has been turned on its head. The relatively streamlined approach of hardware development has turned into more of a scramble to see what components are available today, buy them, then design them in. Design for Production used to consider a wide range of subtle production issues, with the assumption that diligently chosen parts will be available on-demand. This has now turned into the problem of forcing production through using a cache of critical parts, just to have something built rather than nothing. At Hitex, we’ve now had multiple projects where new hardware has been designed and software ported just to move a platform to devices that are currently available.
Striking a balance
It’s time to re-think the oft-thought ‘silly’ scenarios in risk analysis discussions. Those cases where the likelihood is incredibly small that they’re deemed implausible – we’ve had to live through a number of them recently, and the industry is still struggling to recover.
If the option is available, now may the time to develop for when the supply chain tries to return to normal. There may be enough components available now for prototypes to perform a development cycle – ready for volume production as suppliers gain stocks again.
It still effectively comes down to what you are trying to achieve: low-value, low-complexity jobs will always be cheaper in countries with lower overheads, but where the cost of getting something wrong and having multiple iterations of design can make a product unviable, bringing the work home to roost – aka “reshoring” – will pay dividends. Just trying to produce anything right now requires on-the-ball management, all the way from design through to final production.
Contact Details and Archive...