4 Big Data challenges faced by manufacturers

Business impact doesn’t come from having the information, it’s from what you do with it - learn from a light bulb example.

by Marc Leroux

I remember when the US based hardware store Home Depot announced their first light bulb that would plug into an ordinary socket and be controlled through an ios or android app - to be able to turn lights on or off without getting up, or even via the internet when travelling. I was still struggling with the value of having to find my phone to be able to turn lights on, but it was pretty impressive to think of having internet enabled light bulbs. Over time, we got more reasons to switch to a smart bulb - from brightening up the party to improving your sleep. Of course, it would be a lot more beneficial, at least from my perspective, if it also provided useful information such as the remaining life before it needed replacing, but this is certainly indicative of the direction almost every manufacturer is headed, with just about everything.

The light bulb is a great example. If you can buy this at Home Depot, wouldn’t you expect that everything you buy for a manufacturing facility has the same capabilities? In a very few short years we’ve moved from smart devices being “cool” to an expectation that everything is an intelligent device. Fortunately, that’s an area that ABB, along with other automation vendors, has focused on since a long time. By intelligent we mean that the devices need to be able to run, collect data, understand their current status or health, communicate with other systems and devices, and be able to react to configuration or operational changes in a secure fashion. And that leads us to some of the bigger issues when we start talking about the Industrial Internet of Things.


In a very few short years we’ve moved from smart devices being “cool” to an expectation that everything is an intelligent device - able to run, collect data, understand their current status or health, communicate with other systems and devices, and be able to react to configuration or operational changes in a secure fashion.

Share this page

1. Applying proper context to integrated data

Typically a device is part of a larger system that manages a process, and with most processes there needs to be coordination. In order to get value out of intelligent devices we need to ensure that data is transformed into actionable intelligence – data with the right context applied to it and data that includes operational experience.

Having a small set of intelligent devices is pretty reasonable to understand and manage. Many intelligent devices for manufacturing allow parameter adjustment (or control) and most have status or diagnostic information, again pretty easy to manage with a limited number of devices. So perhaps the first challenge we face is that while having autonomous devices is pretty interesting, the reality is that it would be impracticable or even undesirable to have them operate autonomously. Typically a device is part of a larger system that manages a process, and with most processes there needs to be coordination. After all, wasn’t this the reason that manufacturers recognized the need for Control Systems so many years ago? The concerns over security/safety and process optimization/profitability don’t go away because we have intelligent devices. What does happen, though is that we start to look at new ways to incorporate devices. We have been talking about “Islands of Automation” for the past 20 years, and they still exist today, and it is pretty normal to find that a manufacturer may only have 30% of his field devices connected to a Distributed Control System (DCS). One of the reasons for this has been the cost to connect them. With intelligent devices, the promise is that connectivity is built in, reducing the interface cost, but we are still faced with the often overlooked integration issue: unless there is context to the data, then it is just data!

Metcalfe’s Law, formulated circa 1980 postulated that the value of a (telecommunications) network increased exponentially with the number of connections. If you had 2 subscribers with telephones, then you had one possible connection. With 100 subscribers it would be possible to have 10,000 different connections. While that is true, the flaw in the argument is that all have to be speaking the same language. If 50% of the subscribers speak only English while the other 50% speak only Swahili, the number of connections (value) goes down significantly. Even if we agree on a common language, we still have to factor in experience. My Swahili speaking friend may talk about “Spar” and I would be wondering what the discussion has to do with ships, while they would be talking about a retail store. Only if we have a common context (language and experience) do we get the full value of the network.

We have the same issue with intelligent devices. Unless we have some type of model that brings everything to the same context level then the value of making decisions not only decreases, but can become dangerous. Think about what would happen if one device returned a status of “1” that meant “I’m going to fail soon” while your program was expecting that “1” meant “everything is OK”. In order to get value out of these intelligent devices we need to ensure that data is transformed into actionable information, which means that some type of contextual model needs to be applied to all of these intelligent devices.


Want to improve coordination between operations, engineering and business management?

5 value pillars of digitalization

2. Infrastructure to support ever growing data

Big Data requires more than just a “fast network” it’s what that network looks like (wireless/hardwired), what protocols are used, disk storage needs to be considered, display form size and the structure and amount of data retrieved. Without some expert help, it’s not easy to get this right.

The second issue is that manufacturing facilities tend to have a lot of devices, and devices are increasingly generating more data, which leads us to another buzzword that has been around for a few years: Big Data. While most people tend to refer to these as independent concepts, but the reality is that they really need to be looked at jointly. Manufacturers have been dealing with big data for decades, long before the name was coined. Consider that many historians can now collect data at millisecond intervals and there are tens of thousands of measurement points, and we may have been collecting these for 5 years. Isn’t that Big Data? Now we want to collect even more data, from even more devices, making our existing “Big Data” even bigger. This means that we have a few things to consider, primarily about the infrastructure to support this. It’s is more than just a “fast network” it’s what that network looks like (wireless/hardwired), what protocols are used, disk storage needs to be considered, display form size and the structure and amount of data retrieved. Bottom line, without some expert help, it’s not easy to get this right. We tend to think that physical disk storage is becoming less expensive (which it is), but we still need to consider backups, archiving and physical space to be considered. Today, most people want to be able to see data on mobile devices, which may have a smaller footprint than desktop devices, making it more convenient, but harder to visualize. Dealing with Big Data is complex, and most manufacturers are less interested in this than they are in how they can use this data to improve their operations, which leads us to the third significant issue: analytics.

Need expert support with data collection and storage?

Connectivity and data storage solution

3. Software to calculate the impact of decisions

Instead of having an engineer look at a series of trends and try to manually determine the cause of a problem, the expectation is that software will go through process values, alarms, events and observational information to determine the root cause. Once the cause has been determined and validated, the software will update the model to generate an alarm before the situation occurs again.

Several customers I’ve talked too lately have lamented on the fact that they have fewer engineers to evaluate process related problems, the ones they have are younger and less experienced, and now we have much more information that they have to look at. The discussion has typically quickly moved to the need for analytics that will automatically sift through the data and find the root cause of problems, or highlight inconsistencies in the data. Instead of having an engineer look at a series of trends and try to manually determine the cause of a problem, the expectation is that software will go through process values, alarms, events and observational information to determine the root cause. Once the cause has been determined and validated, the software will update the model to generate an alarm before the situation occurs again.

Analytics can also help in other ways. A maintenance manager I talked to told me that he wasn’t interested in having more data, he had too much as it was. He knew he had problems but because he didn’t see them correlated to a bigger picture he ignored most of it. It is not enough to see that there is a potential problem, we need to see what the business impact of making a change is going to be.

In the business environment he was working in the cost of replacing or repairing something had to be balanced against the value of keeping it running, or delaying the repair until high value production orders were fulfilled. This was a problem that he didn’t have a solution for, so he relied on “gut feeling” to make most decisions. Without having software that could calculate the value or impact of his decisions in the context of the rest of the business, having this “Big Data” wasn’t helpful at all.


Looking for software-based solutions with systematic approach to problem-solving and proactive service?

Industrial software

Predictive notification program that focuses on high-value actions

Rules established for the predictive notification program

4. Bringing the IIoT capability to older devices

While this is just touching the topic of IIoT, one of the often overlooked challenges that manufacturers face is that a large percentage of their equipment was purchased before we’d heard of the IIoT, and a large percentage has a lifespan of 10-20 years. For example, three 100 year old ABB transformers were recently retired in Australia, and while they were certainly ahead of their time when they were installed, they didn’t have a lot of connectivity options!

Whenever we look at an IIoT strategy, we need to consider how to provide meaningful, cost effective solutions that bring the IIoT capability to older devices.  At a minimum, these need to be included in the strategy in terms of evaluating the value that they can bring.


Convert traditional equipment into smart, wirelessly connected devices

Smart Sensor

IIoT isn’t just plugging in devices

The Industrial Internet of Things isn’t just plugging in devices. That is the starting point, of course, at least for newer devices. Manufacturers, though, at least the ones that I talk to, are really interested in how they can improve their productivity, profitability and safety, which means that the real emphasis needs to be on what you do with the information. The IIoT includes the collection of raw data, transformation of the data to information, storage, and most important, analysis of the information. The business impact doesn’t come from having information available, it’s from what you do with it. This is the point that sometimes gets overlooked when we talk about the IIoT and Big Data.


The business impact doesn’t come from having information available, it’s from what you do with it to improve productivity, profitability and safety.

Learn more

Augmented Reality in manufacturing
Digital trends reshaping industries
Big Data and decision-making in industrial plants
  • Contact us

    Submit your inquiry and we will contact you

    Contact us
Select region / language