In March 2020, our world was changed forever. A novel coronavirus - COVID-19 - swept across the globe, leaving in it's wake, morbid mortalities, crippling fear and the resurgence of 5G conspiracy theories. As we emerge from the scour of the virus with life-changing scars, I've been reflecting on how the coronavirus might change the way we work as electronic engineers....
Advances in the rest of the world
In some ways, I think the world stood up to the coronavirus. We isolated ourselves in our homes to protect one another from its deadly sting. Our brilliant medical scientists, researchers and doctors collectively raced to develop vaccines four times faster than we had ever previously accomplished. We organised our collaborative work around Zoom and Microsoft Teams, ushering in the age of the widespread use of the phrases "I can't hear you - you're on mute" and "who's the strange fellow in your background?"
Entire industries were transformed, some for better and some for worse. The healthcare industry seems to have implemented years of digital transformation initiatives in months; from the rise of the MRNA vaccine to the proliferation of telemedicine. On the other hand, the aerospace industry ground to a standstill, with airlines wondering when people would next be ready to take to the skies huddled in narrow, metal tube, in groups of a hundred.
How though was the electronic systems industry impacted?
There are many answers to this question, but I want to narrow-in on one that you'll probably already be familiar with: semiconductor shortages and the rise of design for procurement.
When the coronavirus hit, manufacturers, rightly, panicked. Many Automotive, Aerospace and Industrial manufacturers shut down their operations because of the uncertainty surrounding the pandemic. This sent a ripple effect through the supply chain to semiconductor manufacturers who reassigned production from the aforementioned industries, where there was little to no demand, to the consumer electronics and computing industries, where there was an upsurge in demand.
For each individual company, this was the right thing to do, but for the world, this would later cause unmitigated chaos. This is because shutting down the supply chain happens in a flourish - instantly - but starting it back up happens in a drag - slowly and unwieldy.
So here we are, in a world where the most important components in our designs have six month lead times; where components that were available when we started designing at 10:00 am, by evening, will only be available in July 2022; where distributors accept our orders only to let us know later that they can't fulfil it until months in advance.
Engineering teams are resilient
It's not all doom and gloom, though. While some companies have made the decision to slow down or even stop development until this crisis is averted, many cunning, resourceful engineering teams are hacking their way to the new world of designing with procurement at the top of their minds.
Of course, design for manufacturing has always included taking into account component availability and pricing etc. However, I have always found that for most companies this was usually an after-thought. We would perform our designs first, then worry about component availability, preferred suppliers, pricing and supply chain resilience later.
The exception is large OEMs (especially those who perform high volume manufacture) who have entire purchasing teams who maintain special relationships with suppliers to ensure the resilience of the company's supply chain. This came naturally to them as design for procurement was a critical part of their business plan.
For the rest of the electronics design world, we are making our way bravely into this new world, experimenting vigorously with established and novel approaches. I have personally learned a lot and I'd love to share some of the tactics I heard about which I think anyone can try:
1. Component Alternatives
One increasingly common tactic involves selecting components which have a close alternative. For instance, I recently came across a comment from an engineer who designed three footprints for a single component. Although painful, embedded in this approach is the implicit requirement that there be alternatives for the chosen component that have similar electrical connectivity.
2. Categorising Components and Ordering Early
An approach that bundles component alternatives into a more holistic method involves categorising components based on their importance and ordering critical components early.
Category 1 components such as resistors and capacitors, which have a plethora of alternatives, are optimised and procured when the design is completed - maybe even by the manufacturers.
Category 2 components such as LEDs, connectors, basic amplifiers, which are a bit more scarce, have alternatives (usually with similar connectivities) selected and specified in the BOM.
Category 3 components, such as converters, controllers, microprocessors, which are scarce yet critical are chosen and ordered very early in the design phase (maybe even before any ECAD tool has been opened)
3. Parallel Redundant Design
A frighteningly expensive suggestion, which may be reserved for crucial projects within large companies involves some companies choosing to have multiple teams working in parallel to develop motherboards for the same product to ensure functional redundancy and supply chain resilience. This might seem like overkill until you realise how much more Sony could have saved if they employed this method earlier, instead of reportedly needing to re-design the PS5. To be honest, this approach seems like overkill to me too.
What other interesting approaches have you seen?
I'm inclined to start thinking about design for procurement as a different concept from design for manufacturing, if only to sharpen the importance of the discipline during the front end of the design process.
In addition, I've recently been thinking about why these tactics and processes are necessary in the first place. I long for a world where electronics engineers and prolific inventors of new smart devices are unencumbered by routine, tedious issues like component availability.
In fact, at Circuit Mind (funded by investors like Google, ARM , Episode 1 and Join Capital), we are already building the intelligent software that can help engineers take design for procurement into account during design. Let's see how it goes....