Are factories going to adopt automation quickly or nah?

This story was originally published on June 23, 2023 in my weekly IoT newsletter. You can sign up here.

Will factories embrace digital automation rapidly in the next five years, or will they continue to adopt technology in an incremental fashion? That’s the pressing question behind a recent McKinsey report that tries to identify the biggest trends in industrial automation and predict the types of vendors that will gain the most market share as manufacturing companies embrace automation.

While this isn’t exactly the “Clash of the Titans,” it’s a fight I’ve seen coming for the last eight years I’ve spent covering this space. McKinsey lays out four categories of vendor, hyperscalers, automation giants, software providers, and specialized players, and concludes that the hyperscalers are moving the most aggressively and have the most to gain if manufacturers adopt digital tech in the near term.

First let’s talk timing. McKinsey points out that while a few companies have adopted digital automation wholesale through new construction, most factories have added new technology piecemeal. This makes sense given that companies have historically built factories around a process that would operate the same way on the same machines for a few decades.

For the IT crowd, think of a factory as an integrated data center before machines became hot-swappable and the servers themselves were virtualized. Adding new technology was a big lift and a big risk. But thanks to new technology such as sever virtualization, soft PLCs, and robust and resilient wireless networks, factories are gaining some flexibility. They are becoming less hardwired in terms of the physical machinery and the computers that control those machines.

These tech trends, plus economic factors such as the need to compete against those who have already adopted more automation, a lack of qualified employees, and a new focus on sustainability could push those who haven’t fully adopted automation technologies to move more quickly. This means they may rush (in manufacturing terms) to join their more automated competitors within the next five years.

The McKinsey report doesn’t go into this but I think we’ll also see factories that produce higher-value goods adopt technology faster. If the government increases fines for pollution or for manufacturing facilities that produce essential goods but don’t have good cybersecurity, we may also see that drive more factories to adopt new tech more rapidly.

This time frame of five to 10 years is what McKinsey dubs its revolutionary scenario, one that will likely lead to hyperscalers such as Amazon, Microsoft, NTT, etc. winning more of the market for industrial automation at the expense of traditional automation companies such as Honeywell, Emerson, Siemens, Schneider Electric, Rockwell Automation, ABB, etc.

But if the aforementioned technologies — plus others including teachable robotics, better additive manufacturing (3-D printing), matrix production, and better harmonization of data — do not hit inflection points and see their use increase, factory automation may grow more gradually, hitting widespread adoption around 2040.

Briefly, teachable robotics is a mid-term technology where instead of programming robots, engineers could use AI to show robots how to perform a task, making robots easier to use and more flexible. Matrix production is the use of flexible cells on a factory floor for producing parts as opposed to one long, continuous line. Data harmonization is the adoption of standards such as the Open Platform Communications Foundation’s Unified Architecture (OPC UA) to readily exchange factory data between different systems. The report notes that in 2022 there were more than 4 million OPC UA network nodes, up from 285,000 in 2018.

For companies looking at this report and asking which scenario to plan for, they’ll get no answers. But it does provide some obvious paths for both hyperscalers and automation vendors to take, most of which involve things like evaluating their current products aimed at industrial automation and moving further up or down the stack. Both should look to M&A to add more capabilities, which is good news for the smaller startups that are rampant in this world.

Both hyperscalers and automation vendors are also encouraged to acquire or develop through partnerships specialty expertise in specific industries. Actually, both hyperscalers and automation vendors should do that regardless of how fast industrial automation gets adopted, according to McKinsey.

Finally, automation vendors should keep an eye on hyperscalers and in the short term, cultivate data and expertise that they can provide as a service to end customers, and hope to hold on to market share if there is revolutionary adoption. And if that adoption turns out to be more gradual, they can engage with hyperscalers using that specialized expertise as a trusted Tier 1 partner.

In short, companies that are trying to buy tech for industrial automation should get ready to play the hyperscalers against the industrial automation companies, and to use their market power to keep both in check.