Protecting Worker Safety in the Cloud Computing Supply Chain

We must continuously improve our safety management strategies to protect our employees.

Cloud computing sounds so benign. Just the language we use to describe this technique for accessing apparently unlimited amounts of computing power and digital storage over the ether sounds so free of consequence. How could it be anything else, when clouds are so soft, so remote, so transient?

Of course, the cloud-computing infrastructure that enables the online services that we have come to rely upon is anything but virtual. The continuous operation of data centers is so important that they include multiple forms of redundancy, for everything from their Internet connections to their power supplies.

All this infrastructure has to be designed, built, bought, shipped, installed, commissioned, operated, maintained and upgraded – and these are real physical tasks undertaken by real people, not the mythic creatures that the term ‘cloud computing’ might suggest. As we move more of our lives online and the number of data centers in the world steadily grows, it’s down to all of us in the sector - from genset manufacturers to hyperscale operators - to ensure that we recognize and manage the consequences of their use at a personal, corporate and customer level.

This means that, at every point in the chain, processes must be optimized to ensure the optimum level of protection for workers.

At Kohler, for example, we have established a corporate culture of ‘Zero Injury is Possible’ – what we call our ZIP strategy. It guides Kohler’s approach to accident prevention across our global power group sites, and says that no one should be injured in the workplace and that all accidents – however minor – can be avoided.

Strategies and cultures are easy to talk about, harder to implement. We try to deliver on our statement of intent by recognizing that protecting our employees' health and safety is our top job and our most important value. Making this ethos stick requires a top-down approach to safety, with leadership providing the authority and commitment to ensure that our facilities are as safe as they can be.

Having set the culture, our next step is to apply a risk-based, data-driven injury prevention program (IPP) consistently across the shop floor. This program is also underpinned by the ethos of continuous improvement and monitoring of potential risks that are embedded in all our manufacturing processes.

Let’s look at what the IPP means in practice – using one of our main generator production sites in Brest, northwest France, as an example. Here, any potential accident hotspots are identified and analyzed using simulations of people and material flows, to enable plant managers to predict where issues might occur.

Regular leadership-led safety tours are conducted at the plant, with a strong focus on employees’ postures, behaviors, and interactions with their environments.

Meanwhile, a ZIP card process helps encourage associates to flag up any safety issues. Workshop managers and operations leaders regularly review these ZIP cards, and then implement processes and safety measures to drive improvements.

It’s a doctrine that seeks out lots of small safety enhancements, in the belief that together they will add up to large improvements in outcomes.

For example, the ZIP card process revealed that a subcontractor wasn’t using Personal Protective Equipment properly. This was immediately addressed, and then followed up later to ensure that the subcontractor continued to work safely.

Other safety responses have required more significant time and investment. After a near-miss between a warehouse worker and a forklift, new technology was implemented to limit the forklifts' speed and lift function in some areas of the plant.

None of these approaches would work without a culture of openness in which all employees are encouraged to have their say. Stand-up meetings at the start of each shift provide an opportunity to discuss safety concerns. We use the same approach in non-manufacturing locations, where risks and inappropriate safety behaviors need to be discussed and assessed.

As an engineering firm, we like to take actions and then check whether they worked by taking measurements and analyzing the resultant data. We’ve done this by tracking the recordable incident rate at the Brest plant over the past three years. In 2019, the rate was 3.4. In 2020 it was 3.0. And in the last quarter, it was just 1.5. Our safety efforts so far have set our accident rate on a long-term downward trajectory.

That’s fine and good, but with demand for data centers booming we have to anticipate what will happen if we start producing much larger volumes of gensets. It would seem this would create more opportunities for injuries, so it is up to us be even more vigilant about current and potential threats to the safety of our workforce.

It naturally follows that this provides assurance for customers seeking to plug into a responsible supply chain. The news cycle might be increasingly short, but working with a partner that is actively reducing the risk of injury - significant or otherwise - is a sensible step towards avoiding negative press by association.

For all that the language of cloud computing makes the service sound benign, it is important for those of us who provide its infrastructure to recognize that it is not entirely risk-free. It’s up to us to apply continuous improvement doctrines to our safety management strategies, to protect our most valuable resource, our employees. And maybe it is also up to end users of cloud computing to think a little more about how those services are implemented. If we can all focus a little more on the real rather than the virtual, then together we can ensure that every cloud service does, indeed, have a silver lining.