We get two very different types of calls when an industrial PC fails. The difference between these calls is whether someone thought about spares before the failure even occurred!
The first type of call is a calm one. The customer has a spare on site, they have swapped it in, and the plant is running. They are calling because they now have a single point of failure and want to discuss a longer-term solution. There’s time to evaluate alternatives properly, and to reassess the broader system, including software or I/O card upgrades to improve long-term maintainability.
The second call is the urgent one. The plant is down. There’s no spare. The customer can’t wait the standard two to three-week lead time. They phone every supplier looking for something in stock locally, even if it’s not ideal. In some cases, they pull a desktop PC from the office and sit it on the factory floor. We have seen those desktop PCs still running in production years later, slowly failing in conditions they were never designed for.
The logic for holding spares is obvious. but so is the logic against – it ties up capital in hardware that may never be used. This often means the decision gets deferred because there is no clear framework for making it.
In practice, whether you hold spares comes down to how critical the system is, how long a replacement might take and what the downtime would potentially cost.
What stops production?
Not every industrial PC needs a spare. The ones that do are where a failure would stop production, create a safety risk, or trigger regulatory consequences. In most plants, this is a short list: a SCADA server, a line controller or a batch management system.
Systems that are important but not immediately critical, such as monitoring displays or data loggers, can tolerate a longer replacement window. For those, having a documented specification and a reliable supply path matters more than holding physical stock.
How fast can you recover?
Most industrial PCs are built to order. A typical lead time is two to three weeks. Longer if components are on extended lead times or have been discontinued. With the current global shortage affecting RAM, SSDs and CPUs, lead times have been blowing out to months in some cases, with significant cost impacts.
If your critical system fails and you have no spare, two to three weeks is the best case. Longer if the configuration was never documented and needs to be reverse-engineered. For remote sites, add days regardless of how fast the hardware ships.
A simple framework offers a solution
For each critical PC, ask:
- What does downtime cost per day? A rough estimate is enough. If the answer is significant, a spare is justified.
- How long to source a replacement? Factor in hardware lead time, software configuration and installation. If measured in weeks, a spare that swaps in within hours changes the equation fundamentally.
- Can the spare be like-for-like? The simplest approach is to order a spare at the same time as the primary unit, built to the same specification. This is what most of our customers who plan for spares do.
- Is the system documented enough to rebuild without the spare? Do you have the hardware spec, software media, licence keys, and a current disk image? If not, the spare becomes even more critical.
Don’t overlook the importance of disk images
A spare PC is only useful if you can get it running quickly. That means having a current disk image.
This is the step that gets neglected most often. We regularly see situations where a disk image was created at commissioning years ago, but nobody knows where it’s stored. Staff have changed, the image was on a USB drive in someone’s desk, or it’s years out of date.
A disk image should be treated as a critical operational backup, not a one-off commissioning task. Schedule it regularly, store it in a known location and verify it can be restored. The spare PC and the disk image together form your recovery capability. Either one without the other leaves a gap.
What about obsolescence?
The concern that a spare might become obsolete before it’s needed is understandable but rarely a practical problem. You are buying insurance against a high-consequence risk. The cost is a modest premium for operational resilience.
The more serious obsolescence risk is the one that affects your ability to source a replacement at all. Even industrial-grade motherboards, RAM and storage are eventually discontinued despite their longer lifecycles. If you have not secured a spare and the components are no longer available, the recovery path becomes finding an alternative platform and revalidating the entire software environment. That takes significantly longer and costs significantly more.
Your spare hardware strategy doesn’t need to be complicated. Identify the PCs where failure has serious consequences. Order a spare at procurement time. Maintain a current disk image. Keep a record of the hardware specification so your supplier can replicate or recommend alternatives without starting from scratch.
A spare industrial PC costs a few thousand dollars. A day of unplanned downtime at most industrial sites costs many times that.
How ESIS can help
We keep records of every system we build, so when a customer needs a replacement or spare, we can replicate the original without starting from scratch. If components have become obsolete, we work through the requirements in detail and find an alternative that meets the same needs.
If you want to put a practical spares strategy in place, we can help you identify which systems are most at risk and ensure the builds are documented for future reference. Our engineers are always happy to have a chat, whether it’s about technical questions, spares planning or obsolescence.
Contact us to discuss your strategy with our experienced engineers.
