Continuing a discussion from my last column on the Industrial Internet of Things (IoT), let’s move into the process industry realm where I believe there are “fewer fools” because control professionals understand the basic requirements for configuration and testing. That’s not to say we don’t make mistakes, but we have standard change management and related procedures to circumvent accidents. We’re also aware of the potential consequences of mistakes and act accordingly.
So other than IIoT devices being more critical than IoT devices such as your doorbell, what does the extra “I” really mean?
- IIoT devices are widely distributed, so they’re not always in a secure environment (i.e. inside the fence line). And there are additional end points, so in many cases, they’re more vulnerable than traditional field devices.
- Provisioning associated with initially connecting a device to the control system must be part of the security and architecture design.
- Automation with limited or no human intervention allows a system to fail gracefully, and recover from intermittent service interruptions or updates.
- Integration with legacy and different sensors and actuators connects the control system in multiple ways, including to cloud-computing services.
- Unlike devices in a home, most industrial installations deploy hundreds or thousands of devices over wide areas and different applications to support reliable operations and asset management.
- Higher levels of accuracy are available, not only for measurements, but also to handle latency and response times because automated, high-speed machinery is synchronized down to milliseconds.
- Flexibility to be programmed or configured to meet specific applications. A single sensor can measure multiple different ranges over its life, depending on where and how it’s deployed.
- Industrial products are designed to operate for 20-30 years before a scale replacement is performed. Plus, they must run with high availability and reliability since poor data from I/O reduces model accuracy.
- IIoT devices are often deployed in harsh environments that are regularly subjected to extreme cold, heat, vibration, pressure and hazardous/flammable (gas/dust) conditions.
- System architectures must be resilient and always complete their processes, so they don’t negatively impact operations.
- IIoT devices must be serviceable to sustain the levels of performance required over anticipated lifecycles. Serviceability extends from swapping out sensors or power supplies and updating firmware to configuring gateways and servers without impacting other devices on the network.
- Advanced analytics are crucial to extract actionable insights, including self-diagnostics, from the vast amounts of machine-generated data that drives major operational improvements. They do this by enabling predictive maintenance, real-time process optimization and identifying potential bottlenecks.
We tend to take this list for granted because that’s what control systems do. However, home IoT devices are intended to be easy to set up, but aren’t suitable for all the other requirements. Though an IoT sensor may provide the signal, it may not be able to do it for the multiple years needed in industrial settings. Masking an IoT device in another shell to make it look like an IIoT device, or worse, fake, forged or knock-off units that look the same right down to the packaging is another challenge.
One possible way to reduce the likelihood of a fake or forged unit is with digital nameplate or electronic labels that are directly linked to the manufacturer to provide original certificates, user manuals, software updates, etc. Unfortunately, electronic labeling has too many global variations by country, industry or even manufacturer. The recently completed IEC 63365 committee’s work on a digital nameplate standard, known as IEC SC65E WG2, is one such standard for process transmitters, but there are similar standards for medical devices and others. To help with the interoperability of these electronic labels, NEMA has started an initiative analogous to the unique identifier (UID) used in asset management to define a baseline for all intelligent labels regardless of industry or application. Time will tell if they’re successful.
