Smart Heavy Metal: Retrofitting IoT in Brownfield Industrial Sites

Industrial enterprises know they can see significant benefits from the Industrial Internet of Things (IIoT). But they also often have significant investment in existing assets, such as manufacturing machinery. While some may dream of ripping out all of their old equipment and replacing it with completely new IoT-capable machinery, it simply doesn’t make economic sense to dispose of well-functioning capital equipment before the end of its useful life.

But IIoT can improve maintenance, energy use, productivity and product development. In a competitive environment, businesses know they can’t afford to continue to manage their operations as they have.

If you work for one of these enterprises, you’re probably already looking at retrofit solutions, which update existing equipment with IoT capabilities. But you’re also anxious about possible downsides, such as investments that don’t finally deliver improved performance.

There are several important things to consider before jumping into a lot of vendor capability presentations. Failing to take these into account can lead to a lot of wasted time and money.

Understand What You’re After

The essential first step, before making any specific decisions, is to understand the business outcomes or customer needs an IIoT upgrade will help satisfy. Data generated without a clear understanding of its purpose can lead to confusion and analysis paralysis.

What's your highest priority? You may be looking to extend the useful life of your equipment, to avoid unplanned outages, to streamline operations and to increase efficiency; deciding your focus determines a lot of your subsequent decisions.

Don’t keep your thinking within the limits of your current connectivity, but understand it so you know what can be usefully upgraded or reused.

Use What’s Already There, but Take Care

The desire to get data from equipment is certainly not new. Machine-to-machine (M2M) communications via programmable logic controllers (PLCs) and supervisory control and data acquisition (SCADA) systems have been around since the 1960s, providing data logging and telemetry. But these systems and their sensors are meant to operate in a local, closed-loop environment, for minute-by-minute sensing and control. Introducing IIoT expands the types of things sensed, the area over which they can be sensed and the time horizon, by making possible the integration of data over time to show performance.

Any manufacturing facility could have a variety of different M2M networks, acquired or developed at different times, for a variety of purposes, alongside purely manually operated equipment and standalone computer numerical control (CNC) systems controlling such things as drills and lathes.

All of these were intended for local problem solving and did that job well. Now they need to be integrated into a larger, longer-time-horizon IIoT implementation.

You Can Observe a Lot Just by Watching

If existing sensor networks just don’t do what's needed, relatively simple sensors can provide a wealth of information. Because various forms of vibrations in reciprocating motors are diagnostic of various failure modes, analyzing the output of a simple vibration sensor can track and anticipate increasing wear or other problems. Remedial action can range from adding lubricant to taking to the machine offline at a convenient time to replace something in advance of failure.

Fill levels, temperatures, valve settings, whether a door is open or closed...it can be surprising how operationally beneficial even basic sensing of the factory floor can be if continually updated and then analyzed for trends.

The Machine Equivalent of Wearables

Sometimes, instead of modifying existing sensing and control equipment or adding new sensors, you can create a parallel channel for existing data displays.

Some products can take the outputs of gauges and other displays and convert them into digital data, something that can go as far as visually tracking the position of the needle on a dial with a clamped-on reader, turning an old analog device into a digital output.

While that can seem almost absurd, it can be significantly cheaper than shutting down a facility and replacing all of those gauges. A large nuclear power plant, for example, might have hundreds of gauges, and shutting down operations to upgrade them can be a challenging, and risky, proposition.

With PLCs and SCADA, the original system can continue to send or display its data as it always did, while a new parallel channel pulls that same data into the IIoT infrastructure.

Internet of Things Industrial Applications - Understand Your Options

Retrofitting is getting more straightforward, and there are an increasing range of choices. More and more processing is moving to edge computing as sensors themselves increase in processing capability. Combinations of sensors can come with predictive and video analytics built-in.

The gateways that combine sensor data are also becoming more capable and now often translate a wide range of protocols and provide a variety of high-speed direct connections between equipment and software.

Just remember, you're likely to be upgrading and replacing equipment over time. Take that into account when you lay out your network.