TSMC Foundry ensures that all staff in its Taiwanese factory are wearing masks in common areas

Supply chain issues in a world of Covid-19

Image credit: Taiwan Semiconductor Manufacturing Co Ltd

Ongoing Covid-19 disruption to the global supply chain on which high technology depends is self-evident, but the statistics still make for sobering reading and underline a necessary call-to-action.

Heathrow Airport is the UK’s largest port by value, with 34 per cent of the country’s cargo passing through and 95 per cent usually carried in the belly-hold of passenger planes. However, since mid-March it has seen a sharp rise in cargo-only flights from roughly 50 a month to, at time of writing, 50 a day.

Meanwhile for Dover, Europe’s busiest ferry port which handles 17 per cent of the UK’s trade-in goods, quarterly statistics are due soon but are expected to show a sharp decline. For comparison, at the US Port of Los Angeles, a key entry point for trade between the US and Asia, operation is already down by a fifth.

Activity in much of Asia is beginning to return to earlier levels but slowly – and cautiously. While China is looking to re-ramp output as the world’s largest volume manufacturer, the process is necessarily gradual. With internal travel restrictions only now being relaxed – particularly around Wuhan, both the centre of the coronavirus outbreak and one of the country’s main industrial hubs – itinerant staff are just now returning to factories and offices.

Significantly, China’s lockdown coincided with the Lunar New Year holiday, the world’s largest annual movement of people as workers return home from economic powerhouses such as Shenzhen, Guangzhou, Shanghai and Beijing. Millions have had to remain there since restrictions were imposed or have done so by choice.

All of this is exacerbated by the fact that logistics generally is no less immune to Covid-19 than any other sector. Even where goods are ready for delivery, there may not be the staff to handle them as they suffer from the disease themselves or – in the absence of widely-available tests in either the UK or US – self-isolate in enforced ignorance.

For high technology, there are two notable ironies in this. According to the latest data from McKinsey and Company, parts of the sector maintain comparatively high inventories. With anything between 130 and 200 days of stock within the chain, semiconductors rank second only to pharmaceuticals. This is probably a result of the use of advanced manufacturing and earlier experiences with sudden fluctuations in demand. However, the numbers fall sharply for the some of the markets taking silicon’s end-products, such as consumer and automotive.

The second irony is that high technology is already a pioneer in fields such as Industry 4.0, digital twins, robotics and lights-out manufacturing, all of which allow it to run factories that are less vulnerable to such disruptions. Design has been a major beneficiary of cloud-based tools, which enable home-working (notwithstanding concerns over IP security) and spread efforts geographically, a particular benefit as Covid-19 centres move from country to country.

As one senior semiconductor executive says, this efforts have so far been largely internal: “Once you are inside the factory, the processes are increasingly automated and you can see a lot [of data from] across the different key suppliers in the supply chain. But, even though you have things like delivery trackers, the situation at the factory gate is not as advanced.”

He stresses that this is not to point a finger at the logistics sector. “It has been a lot easier for us to remove the ‘human factor’. We have more controlled, even hermetic environments. To their credit, the logistics business is investing a lot of money and effort into areas such as autonomous fleet platooning for trucks but they simply aren’t ready to deploy at scale. And there’s been massive investment in software – that’s margin for the FedExes and Amazons and all these companies. But many of these things simply aren’t mature yet and, of course, much of that is as much down to the rate of innovation in our own businesses as anything else.

“Could the supply chain infrastructure be improved? Yes – and it will be. But could anyone reasonably demand that happens immediately and on a massive level? Absolutely not. And even if it could, we [high tech] are hardly the priority for deploying it, except for medical – things like ventilators. The focus for any improvements have to be medical supplies and food and anything else that is critical to keeping heads above water.”

Others echo that last point, making it a priority to offer – as has so much of global industry – resources that can be redirected to priority healthcare requirements. These include design skills and component inventories that can be repurposed as well as manufacturing capacity.

The harsh reality is that while most leading companies have crisis management strategies in the drawer, this coronavirus is the mother of all truly global outliers.

“Though service leaders are familiar with business continuity and disaster recovery planning, pandemic planning is very different because of its wider scope and the uncertainty of impact,” notes John Quaglietta, senior director analyst at Gartner’s Customer Service and Support Practice. “The global and dynamic impact of Covid-19 requires planning for longer recovery times and many scenarios because pandemic events are so fluid, and can change quickly without notice.”

Stuart Carlow, chief research office at ABI Research, is blunter, describing the crisis as “a rough sea for all boats. There will be wrecks and we should be prepared for that.” He adds: “Before any change occurs, there will be a retrenchment in outlooks and a reduced investment in modernisation, as survival instincts trump the drive to prosperity.”

However, the industry is not entirely helpless, though the cull might well be Darwinian.

McKinsey’s consultants have gathered spokes of what it sees as a necessary response under a six-pronged umbrella, in terms of ways to improve the supply chain.

  1. Create transparency on multitier supply chains, establishing a list of critical components, determining the origin of supply, and identifying alternative sources.
  2. Estimate available inventory along the value chain – including spare parts and after-sales stock – for use as a bridge to keep production running and enable delivery to customers.
  3. Assess realistic final-customer demand and respond to (or, where possible, contain) shortage-buying behaviour of customers.
  4. Optimise production and distribution capacity to ensure employee safety, such as by supplying personal protective equipment and engaging with communication teams to share infection-risk levels and work-from-home options. These steps will enable leaders to understand current and projected capacity levels in both workforce and materials.
  5. Identify and secure logistics capacity, estimating capacity and accelerating, where possible, and being flexible on transportation mode, when required.
  6. Manage cash and net working capital by running stress tests to understand where supply-chain issues will start to cause a financial impact.

As words, this programme contains a great deal of common sense. But it will be hard to execute, especially for the many companies who have yet to start on any of the points.

Consider the complexities of supply chain visibility. In the strictest sense, this requires visibility not only into your own immediate suppliers, but also their suppliers as well as alternatives. To get some idea of the scale, one often-cited example is a Japanese semiconductor company that undertook comprehensive supply chain mapping after the 2011 earthquake and tsunami. It allocated 100 staff to the task, but it still took a year to complete.

This process has improved. A clutch of start-ups today offer tools that automate much of the work. Companies such as Elementum, Llamasoft and Resilinc enable mapping down to the raw materials level. Elementum even claims that it can get a company’s mapping ‘war room’ configured in less than an hour.

Take up of these tools has been slow and now the crisis is upon us.

Co-authoring a recent article for Harvard Business Review, Resilinc CEO Bindiya Vakil observed: “Seventy per cent of 300 respondents to a survey conducted... in late January and early February, immediately following the Covid-19 outbreak in China, said they were still in data collection and assessment mode, manually trying to identify which of their suppliers had a site in the specific locked-down regions of China.”

China supply chain graph

Image credit: E&T

There is then the question of finding second sources. Many companies have been looking beyond China – at the peak of its outbreak, customers saw cargo delays of between four and six weeks. There is lingering concern that the country could still suffer a second wave of Covid-19 transmission while its ‘reopening’ is still at an early stage. What about the alternatives?

India was cited as one potential alternative. Then, on March 24, Prime Minister Narendra Modi put his country on a total initial three-week lockdown at just four hours’ notice. Initially, Modi attracted praise for acting decisively in ways other world leaders had not – but as TV pictures spread within hours of huge crowds of itinerant workers struggling to get home, the impression turned rather to one of chaos. The lockdown had been announced as sharply as China’s around Wuhan: the execution looked far less effective.

One result for this article was that an interviewee emailed within hours, asking that her earlier positive comments on India now not be published. An illustration, as if one more were necessary, of the fluidity Gartner’s Quaglietta warned about above. What looks right today, may not be so tomorrow.

(As a side point, she is not identified for the same reason. Right now, there is still a problem that goes beyond supply chain issues with running ‘gotchas’ about people in circumstances where things are so unpredictable and the likelihood that people will get things wrong is so high. We need to be able to admit to our mistakes.)

Moving to the end of McKinsey’s list, there is the question of cash. Capital and liquidity are not normally part of supply-chain discussions. In the UK, Chancellor Rishi Sunak has taken big steps to ensure companies can secure bridging loans, but given the extent of globalisation in so many national sectors like high technology, one question that has arisen is what happens where companies consider using the facility as protection against issues at mostly foreign customers (and in some cases, suppliers) in countries where economic intervention has been less generous.

Sunak has been open about not being able to provide a universal financial panacea. Though he faces demands of more for the UK’s self-employed, this does appear to be one of those cases where he cannot realistically go further. It is a delicate judgement call for the business owner.

The point of everything above though is not to downplay McKinsey’s proposals but to make clear that they will involve a lot of work and careful thinking, and are fundamentally necessary. Consider the following example of where they can be shown to likely be working, here specifically in terms of optimising production.

On 18 March, TSMC, the world’s largest chipmaker, reported that an employee had tested positive for Covid-19. The company immediately placed 30 others who had had close contact on 14-day home-isolation and disinfected relevant areas of its facility.

“This employee is receiving appropriate care at a hospital,” TSMC said. “The company has decided to begin operating in segregated teams. All employees in Taiwan are now required to wear masks in common areas and when participating in meetings or training.”

The company said that it did not expect the incident to affect its operations, which have also benefitted from Taiwan’s widely praised national response to the outbreak. As this article was written, with the two-week quarantine ending, that guidance remained in place.

Of course, you do need the masks and the tests to be widely available to do that. In the UK, that issue seems to be beyond industry’s control for now (and the NHS does need to be the clear priority followed by critical infrastructure workers). But planning and implementation in those areas where it is possible will mitigate the impact of the outbreak. Things do largely come down to blood, sweat and tears.

It is worth considering whether or not staff should be directed to those tasks rather than furloughed, particularly since, first, many can be accomplished from home and, second, it is clear that many of the results will probably need to be rethought and recast.

Wash your hands for health, but don’t wash your hands of your supply chain.


The design chain

Design was a process often seen as coming before and separate from supply chain considerations, though increasingly less so. So, to a greater degree, is standards-setting. The roles played by both, however, have been highlighted during the Covid-19 outbreak.

For design, one clear example comes from the drive to manufacture the ventilators needed for critical coronavirus patients. The OxVent team, led by researchers at Oxford University and King’s College London, has rapidly prototyped a design based on supplementing off-the-shelf components and bespoke parts that can nevertheless be produced on 3D printers, integrating the efforts of engineers, clinicians and manufacturers.

Many other teams are adopting the same approach. This needs to be the quickest type of new product introduction, not only as case numbers rise locally but also so that open-source designs can be shared globally in the expectation that the same parts available in, say, the UK will also be in stock elsewhere. We are all in it together.

All these teams are doing fantastic work, but what about standards?

Here, things are more problematic and particularly for 5G, a technology on which engineering is pinning many of its hopes for future near-term growth.

On 23 March, 3GPP, the body agreeing the 5G specification, announced that it was delaying the ‘freezes’ on two releases within the specification, Rel-16 and Rel-17, by three months. As well as standalone networks, these cover industrial applications in the Internet of Things and autonomous driving that are expected to provide new revenue streams for operators.

The delays are directly attributable to the restrictions on global travel as a result of the Covid-19 outbreak. Many industry gatherings such as corporate developer gatherings and tentpole events like the Game Developers Conference have moved or plan to move online. By contrast, standards discussions involve so many nuances and so much politics that while initially discussions can take place virtually, the process needs to be frozen face-to-face.

The delay is a further major blow to 5G. Just as the first consumer handsets have started to come to market, according to ABI Research, “the sector has been hit hardest [in high technology] by delayed shipments and a weakened development of next-generation products”. It is forecasting a 30 per cent fall in handset-making for the first half of this year.

Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.

Recent articles