© 2023 Strypes Ltd. All Rights Reserved. Please note that we use Disclaimer | Privacy Statement | Whistleblower policy | Cookie Policy.
Many decades ago, the hottest computer technology was a videogame called “Space Invaders”. But the first time you played, it was a frustrating experience.
When trying to zap those 8-bit aliens, you missed constantly. You’d fire at them – but by the time the bolt reached them, they’d moved out of the way. With practice, you learned the success strategy: you won not by shooting at where the aliens were … but at where they were going to be.
In other words, you had to think ahead. Judge. Plan. But when you put your plan into effect, screenful after screenful of alien freaks would fall to your zap-gun. Expert players could keep going for hours.
That’s the approach of predictive maintenance – not looking for problems on your network infrastructure today, but for the problems it’s likely to develop in the future. And it’s the subject of this article – the second in our mini-series that started with the one where we discovered why predictive maintenance wins the game over the reactive and preventive approaches.
Let’s deep-dive predictive maintenance (PdM) and see the concrete benefits to your business of managing your technology this way.
1. A reminder: levels of maintenance
First, a recap. Predictive isn’t the only type of maintenance – it shares space with reactive and preventive.
Reactive maintenance is the easiest, yet riskiest. You solve problems as they come up, responding to the trouble ticket raised on Floor 15 or the server fizzing sparks in the datacenter. Needless to say, this running-to-failure approach carries big risks – because by definition, you’re unprepared for business continuity. It’s not really “maintaining” at all; more like fighting fires.
A preventive maintenance strategy raises the game. Using data like manufacturers’ MTBF figures, you service and replace devices and components while they’re well within their expected operating lifetimes. This is great for business continuity, since failures due to wear and tear are rare. But it’s less good for the bottom line. Because manufacturers’ data is conservative. What if that $40,000 disk array you’re sending to the scrapheap actually had four years of useful life left, given your clean operating conditions? You’ve reduced risk – but at great cost.
And that’s where predictive maintenance makes a difference.
2. Why predictive maintenance wins the game
Predictive maintenance is designed to keep all your technology assets functioning over their optimum operating lifespan – and not just functioning, but functioning in a way that means the most to the bottom line. And that involves looking at more than just the manual. A predictive approach combines:
The use case: what is it used for?
A workhorse PC doing Word documents may stay useful for many years – there are plenty of eight-year-old computers on desks. But for a games developer or commodities trader who needs the latest, fastest hardware to run her cutting-edge algorithms? Useful life might be only a year before the next whizzy GPU is needed. The ROI equation looks very different from use case to use case.
The operating conditions: where is it sited?
Next, consider two datacenters. Or two server racks. Or even two laptops on two desks. The products might be identical, purchased on the same day. But what if the first machine sits in an air-filtered, climate-controlled office building, and the second on a factory floor covered in dust and grime? One setup might be fine for five years; the other will fail in months.
Business criticality: how vital is its role?
One low-level worker’s laptop failing is a pain, but it’s not mission-critical. But what if it’s the CEO’s laptop, and he’s about to give a $100m investor presentation? Or what if it’s a server rack providing cloud apps to 10,000 people worldwide? Perhaps the average worker can survive 24 hours of downtime. But your vital infrastructure can’t afford even one second.
So predictive maintenance wins because it doesn’t just “keep the machines running” – it matches the maintenance need to the best business outcome. You’re interested in what’s best for the business, taking costs, data, and ROI into account.
3. Real data for a real world: sensing and analyzing
Predictive maintenance is based on a real-world model, getting the most out of every device and piece of equipment. It recognizes published failure rates and other manufacturer numbers are just averages; what’s needed are specifics. And that’s the second idea in good PdM practice: it makes use of real-world data-driven predictive maintenance.
Predicting outcomes ahead of time
A reactive approach will wait for a device to fail. A preventive approach will book an upgrade on a set date. But a predictive approach gathers data from sensors and metrics, in real time, and works out – based on analytical models – how long that device can perform optimally. And sets the maintenance date based on that figure.
Perhaps it’s six months; perhaps it’s three years. Perhaps the maintenance event will be a like-for-like replacement, or just a refurbished hard disk.
By combining this data with analysis software, this approach can be very accurate. Many enterprise-scale PdM efforts use AI methods like deep learning, neural networks, and stochastic modeling to forecast what will happen to your infrastructure, when … with enough detail to tailor maintenance plans to each individual device.
How PdM increases ROI
Whatever the planned maintenance event, in PdM it’s always based on what the data predicts, not guesswork. And that’s how PdM boosts ROI, across the whole company. Because by following the data, you’re getting the most performance out of every machine, without risking a breach of business continuity by working it too hard.
4. Getting started: developing a PdM mindset
Moving from a reactive model – or even a preventive one – can be tough for some IT maintenance teams. They may be used to solving problems as they occur, not predicting problems that may be years away. So the first step in adopting PdM is to adopt a different way of thinking.
You’re not “fixing equipment” or “keeping things running”. By forecasting patterns of equipment downtime and matching maintenance costs to critical business needs, you’re optimizing your entire IT infrastructure – ensuring business continuity and contributing to financial outcomes. Which aligns the IT team’s goals with the goals of the broader business. (It’s great for the job prospects of IT staff who adopt this way of thinking, too.)
5. How PdM help with industry challenges
The short answer to the question “Why do PdM?” is simple: because your competitors are doing it. Look at any company whose technology you admire, and you’ll probably find predictive maintenance is part of their processes. Because in our technology-driven world, it affects every part of the value chain.
The front end: maintaining customer satisfaction
Customers today expect 24/7/365 service. Even an hour’s downtime in your Service Center, or a minute’s wait at your website, loses you custom and cheapens your brand. PdM gives you a far greater chance of five-nines uptime, all the time.
The front office: maintaining worker productivity
People are pricey. What’s the cost per hour of all the people who use your cloud applications to get their work done? An unexpected failure in the datacenter shifts that amount from a profitable investment to a hard loss. PdM mitigates that risk and keeps vital infrastructure humming.
The IT investment: maintaining asset usefulness
Disposing of or replacing expensive hardware before its time is bad for business – and bad for the environment. PdM makes sure every piece of equipment performs at its best, for the right amount of time – and is only replaced or upgraded when its job is done. Total Cost of Ownership is minimized while useful lifespan is maximized; it’s another benefit of predictive maintenance.
The C-suite: optimizing competitive positioning
A cost-efficient business is a cost-effective one. Nothing makes a CFO happier than seeing IT investment dollars working hard. And because the CIO can talk the CFO’s language – using data models and predictive analytics to explain precisely why IT investment is needed and where it should be directed – it creates greater understanding about how the IT function contributes to the company’s success.
By now, you may be wondering how to implement PdM in your business. And that’s why we’d like to talk to you.
CONCLUSION: Predictive maintenance is productive maintenance
Longer uptimes. Lower risks. An optimized cost model. Smoother business continuity. There’s just no downside (or downtime!) to adopting PdM. It’s an approach we use at Strypes to manage thousands of devices in a huge variety of businesses, using our Nearsurance approach to combine the benefits of in-person teamwork with the advantages of outsourced software development.
Why not get in contact today to see how PdM could transform your approach to IT maintenance?
Remote Еquipment Мonitoring and Diagnostics for the Agriculture Industry
Download our Success Story to learn more about our way of doing it.
Contents
Add a header to begin generating the table of contents