Waites Sensor Technologies

1. Predictive maintenance has evolved significantly over the past decade. How do you see it reshaping maintenance strategies in modern pulp and paper mills compared to traditional preventive or reactive approaches?

When I started out, maintenance was mostly two things: fix it when it breaks, or fix it on a calendar whether it needs it or not. Reactive maintenance is the expensive one. You’re paying for the repair, the lost production, the overtime, and usually a rush order on parts. Preventive maintenance is better, but it’s still guesswork. Half the time you’re pulling apart a perfectly good pump because the schedule says so; the other half, something fails two weeks before its PM was due.

Predictive maintenance gets you out of that cycle. Instead of guessing, you’re watching the equipment continuously and acting on what the condition data is actually telling you. The AI system learns what “normal” looks like for each machine, catches the early signs when something starts drifting, and gives teams a heads-up while they still have time to plan. For a mill, that means fewer surprises, better uptime, and a lot fewer 2 a.m. emergency phone calls. Honestly, the peace of mind is a big part of it, when you’re not constantly worried about unplanned downtime.

2. Which assets or production processes within a pulp and paper facility benefit the most from predictive maintenance, and why are these areas particularly critical?

Mills have always run on thin margins, so every hour of uptime matters. The biggest wins usually come from anything that rotates or needs lubrication, because vibration and lube condition tell you a lot about what’s going on inside a machine. Motors, gearboxes, pumps, fans, blowers, conveyors, even HVAC systems. Once you start looking, there’s hardly a part of the process where you can’t put a sensor.

There’s a safety piece to this too. Anyone who’s worked in a mill knows some of these areas are hot, loud, and full of chemicals. If a sensor can tell me a bearing is fine, that’s one less trip a tech has to make into a nasty spot just to check on it. Think about all the hard-to-reach spots in a mill and being able to keep an eye on every critical component.

3. How are technologies such as Industrial IoT (IIoT), AI, machine learning, and advanced condition monitoring changing the way mills identify and prevent equipment failures?

The big shift is that you’re no longer limited to what you catch on your inspection route. A lot can go wrong between rounds, and traditionally you’d never know until it showed up as a failure. With IIoT sensors feeding data into AI and machine learning, you’re watching full-spectrum vibration, temperature, and other key readings around the clock. When something starts trending the wrong way, the system flags it and gets it in front of someone who can act on it. We’re now catching problems earlier, while they’re still small, not after they’ve taken a line down.

4. What are the biggest operational or organizational challenges companies face when implementing predictive maintenance programs, and how can they overcome them?

Getting sensors on equipment is the easy part. The hard part is what happens after. One machine can throw off thousands of data points a day, and nobody on a busy maintenance crew has time to sort through that. I’ve seen teams get buried in it. One reliability engineer told us it might take 32 weeks to analyze all the data the team gets in one day.

What makes the difference is having real people in the loop. The AI does a lot of the heavy lifting to detect anomalies, but you want certified vibration analysts (ISO CAT II–IV) looking at what the system flags, figuring out what actually matters, and turning the alert into an action for the crew. Without that, and without a partner who knows what they’re doing and understands these operating environments, you end up with a lot of expensive data and not much to show for it. I’ve worked in pulp and paper mills and bring that operational experience to the collaboration when implementing these programs. 

5. Data is often described as the foundation of predictive maintenance. What types of operational data provide the most valuable insights, and how can organizations ensure its accuracy and reliability?

Data really is the foundation, and the good news is it gets better the more you use it. Every time an analyst reviews an alert and acts on it, the system learns from that. Over time it gets smarter about which readings point to a real problem and which ones are just a machine working hard.

In pulp and paper, the combination that works best in my experience is vibration, ultrasonic, and lubrication monitoring. With those three, you usually have a pretty clear picture of when a piece of equipment needs attention.

Vibration analysis— which we pair with a wireless sensor network—doesn’t work just because the sensor has the highest sample rate, frequency, or resolution. All the bells and whistles mean nothing if you don’t have the right machine information and configure the settings to get good data. If you leave everything at the default high frequency settings, you end up with low-resolution data, which makes targeted diagnostics really difficult.

6. Many pulp and paper manufacturers operate with aging equipment alongside modern digital systems. How can predictive maintenance be successfully integrated into legacy infrastructure without requiring extensive capital investment?

Most mills I know are running equipment that’s older than some of the people operating it, and that’s fine. You don’t need to rip anything out. Bolting sensors onto legacy equipment is one of the simplest ways to bring it into a modern monitoring program. A good vendor will know how to do that on older machines, and once the sensors are on, that equipment feeds into the same system that’s analyzing everything else and telling you what needs work.

7. How has predictive maintenance influenced key performance indicators such as equipment availability, unplanned downtime, production efficiency, maintenance costs, and overall equipment effectiveness (OEE)?

The obvious one is unplanned downtime. When you catch a problem early and fix it on your terms, it doesn’t turn into an outage. But the cost side is just as crucial. You’re only working on equipment that actually needs it, and you’re not paying for overtime, weekend callouts, or overnight-shipped parts because something failed on a Friday afternoon. Add all that up and it’s a serious savings, and your machines run better for longer in the meantime.

8. Can you share an example or case study where predictive maintenance delivered measurable improvements in reliability, productivity, or cost savings within a pulp and paper operation?

Domtar’s Kingsport Mill in Tennessee is a great example. Domtar is one of the largest paper, packaging, and forest products companies in North America, and Kingsport puts out up to 600,000 short tons of containerboard a year. So, downtime isn't a small thing.

When they converted Kingsport into a modern packaging mill, they already had continuous monitoring in place. They could see what their machines were doing. The problem was turning that information into something their crews could trust and actually act on. That’s a common gap we see until someone on the reliability side really owns it. That’s where Waites came in. Within 18 months, Domtar saw its mechanical and electrical downtime drop from around 10% to roughly 3%. 

The way it works is pretty straightforward. Waites sensors are on Domtar’s equipment, Waites analysts review the data, and they work alongside maintenance and reliability folks to confirm issues, decide what gets fixed first, and keep improving the approach. Domtar even set up what they call “The Reliability War Room,” with a 65-inch Waites dashboard on the wall. Their lead vibration analyst and their ultrasonic and lubrication specialist work out of there, catching things and building action plans before alerts turn into failures. It’s become one of the most mature AI-driven condition monitoring programs in the industry.

9. As sustainability becomes a strategic priority, what role does predictive maintenance play in reducing energy consumption, minimizing waste, extending equipment lifespan, and supporting environmental goals?

Equipment that’s well maintained just lasts longer. When you catch a misalignment or a lube problem early and take care of it, the machine runs smoother and doesn’t wear itself out. A machine that’s running right also uses less energy than one that’s fighting itself, and you’re not scrapping and replacing equipment before its time. All of that lines up pretty naturally with energy and waste reduction goals.

10. How important is collaboration between maintenance teams, operations personnel, and data analysts in ensuring the success of predictive maintenance initiatives?

It’s everything. Sensors and AI are great, but none of it matters if the insight doesn’t make it to the tech on the floor in a form they can use. When operations, maintenance, and analysts are actually talking, alerts have the right context: how urgent it is, what’s probably causing the issue, and what to do next and when. Without that guidance, techs just get a long list of flags and a dashboard full of red, and it’s hard to know where to start.

Back-and-forth interactions between vibration analysts and facility maintenance and reliability teams allow us to help with troubleshooting and get to the root of the problem and the fix. A reliability partner like Waites is there every step of the way, answering questions and following up with more analysis. We all want the same thing: make sure nothing breaks.

11. Cybersecurity and data governance are becoming increasingly important in connected manufacturing environments. What considerations should pulp and paper companies keep in mind when deploying predictive maintenance solutions?

Your IT team is going to ask about this, and they should. Look for a partner that built security in from the start, not one that bolted it on later. The best setup is a system that runs completely separate from your production and corporate networks. It should be an independent system that doesn’t need Wi-Fi, an internet connection, or firewall changes, so if someone does get into your IT network, they can’t use the monitoring system to move around. Sensor data should also be stored locally on dedicated, secure receivers. That way you get the monitoring without opening up new holes. That’s how the Waites condition monitoring system is built; it runs on its own private wireless mesh network.

12. How do you see emerging technologies such as digital twins, edge computing, autonomous maintenance, and generative AI further enhancing predictive maintenance capabilities over the next five years?

AI keeps getting better, and that means fewer emergencies on the floor. Waites’ AI has a learning loop built in, so it figures out which sensors tend to throw false alarms and which ones mean business. That makes it a lot better at only bugging your team when it really matters.

The next step after predictive is prescriptive maintenance. Predictive tells you something’s wrong. Prescriptive tells you what to do about it, what to prioritize, and when. Right alert, right time, right fix. And those recommendations are only as good as the people behind them. I’ve spent decades in these mills, and that kind of hands-on knowledge informs the prescriptive guidance.

Down the road, I expect to see more sensor coverage, better integration with the other systems mills already run, and machine learning that gets really good at seeing problems coming.

13. For organizations that are just beginning their predictive maintenance journey, what practical steps or best practices would you recommend to ensure a successful and scalable implementation?

First, don’t panic when the data starts rolling in. It’s a lot at first, and that’s normal. Lean on your reliability experts and build a real working relationship with them, so you can figure out together how to prioritize actions. Cut through the noise. Once you’ve got a game plan for responding to alerts, the scheduling gets easier, and you’ll start seeing the payoff pretty quickly. We typically see ROI in three to six months.

14. Looking ahead, what major trends or innovations do you believe will define the future of predictive maintenance in the pulp and paper industry, and how should manufacturers prepare to stay competitive?

Margins are getting tighter and experienced people are getting harder to find. That’s just where the industry is. Predictive maintenance is going to be one of the things that separates the mills that run smoothly, keeping downtime in check, from the ones that are constantly putting out fires. As the AI gets better at learning from analysts, I think adoption is going to pick up fast. There will always be people in the loop, and there should be. But predictive and prescriptive maintenance are going to be an important part of how mills deal with a labor shortage, because they help a smaller crew focus on the work that actually matters. And it also comes back to peace of mind. With fewer emergencies, workers get more time back.