Robots kept running: what the pandemic tells us about automation
While factories shut down due to COVID-19 outbreaks among workers, automated lines kept operating. What this says about the case for robotics.
In the first half of 2020, several large manufacturing facilities shut down not because of equipment problems or supply chain failures, but because of COVID-19 outbreaks among workers. Some plants closed for weeks. This was a new experience for industries accustomed to treating people as the most reliable element of the production chain.
Against that backdrop, a different pattern became visible: automated lines, where the share of manual labour was minimal, kept running. Not without difficulty - logistics, maintenance, and setup still required people. But the production process did not stop.
That changed the conversation about automation.
How the argument was built before
Before 2020, the main argument for robotics was usually economics: the cost of a robot versus the cost of a worker over several years. Quality, speed, and three-shift operation without overtime were sometimes added.
The resilience argument - robustness against unforeseen disruptions - was rarely made. It seemed theoretical.
The pandemic made it concrete.
What 2020 showed
The vulnerability of facilities that depend on dense concentrations of workers turned out to be real and measurable. Meat processing, assembly lines, packaging - industries with a high share of manual labour faced shutdowns that had not been planned in any risk scenario.
This does not mean everything should have been automated in advance. It means a new parameter appeared in the risk model that had previously been ignored: dependence on people being physically present in a specific place.
How this changes the calculation
Traditional automation ROI analysis counts direct costs and direct benefits. Adding a resilience factor is harder - it is not linear and not regular.
But the question can be framed differently: what did a two-week production shutdown actually cost? If that scenario now seems real rather than hypothetical, it belongs in the calculation.
This is not an argument for "automate everything immediately". It is an argument for "include the cost of vulnerability when evaluating automation projects".
What to consider now
Not all tasks are equally easy or equally justified to automate. A few filters:
- Which of our operations stop if some of our people are unavailable? Those have the highest vulnerability.
- Among those - which are technically feasible to automate at the current state of robotics?
- What is the payback period, including not only direct savings but also the cost of a shutdown?
- Do we have the reliability and failure data needed to maintain an automated system?
The pandemic did not change the technology. It changed how we assess the risks of depending on people being present in a specific place. That argument is now in the working spreadsheet, not in the theoretical section.