Key highlights
- The article emphasizes that modern IIoT implementations are no longer limited to conventional Internet Protocol (IP) setups.
- It highlights that effective IIoT deployment must include robust cybersecurity practices, such as segmentation, firewalls and adherence to standards like ISA/IEC 62443.
Because the Industrial Internet of Things (IIoT) has broken free of its Internet-protocol (IP) usual constraints and definitions, users can apply it or whatever networking combination serves best where it’s needed. Just don’t forget about the sensor, instrument and device level. Oh and definitely don’t forget to apply sufficient cybersecurity, too.
“We think of IIoT as using approved devices to get real-time data, so we can perform predictive analytics that will enable maintenance, optimize our processes, and reduce downtime,” says Chad Paxson, process control analyst at the Cobb County-Marietta Water Authority. “We also use IIoT to enable edge computing and planned artificial intelligence (AI)-aided analyses.”
The utility is the state’s largest producer of drinking water. Its two main plants have a capacity of 173 million gallons per day (mgd), though it usually only needs to produce half of that to meet average daily demand of 79.8 mgd from its nine wholesale customers, including five counties that buy its water (Figure 1). The utility runs 10,000–15,000 data tags at each plant, and operates about 30 primary PLCs, which will be reduced as it implements hot standby, control and remote PLC zones in cyber-secure network segments.
To gain more benefits from IIoT, Paxson reports the water authority began working with Texas-based Specific Energy about a year ago, and implemented its software and edge hardware, which can relay data to a cloud-computing service. They also deployed network segmentation and firewalls for greater cybersecurity.
“We send flow data and pump curves, and Specific Energy’s solution lets us select pumps, performance levels, times and energy use,” explains Paxson. “It’s software and edge device tie to our flowmeters and SCADA information, and link with Georgia Power’s rate calculator. We also recently switched our HMIs to Inductive Automation’s Ignition web-based SCADA software. These tools let us monitor pump performance, wear and tear, determine if running two pumps is better than running three pumps, and seek the most cost-effective combinations and schedules.” (Figure 2)
AI, face-time and clean data
Possibly using some savings from its electricity rate-shaving, Paxson reports that Cobb County-Marietta is also investigating how AI can help it achieve further efficiencies, such as improving leak detection.
“We already check for leaks by monitoring flow and pressure at remote sites throughout 190 miles of pipeline in the county, and we use some modeling software. However, it’s mainly a DIY solution that relies on manual entry of remote SCADA data into Excel to produce trends for the past two weeks,” adds Paxson. “This data also goes to Badger/Syrinix Radar’s website that helps monitor the 14-16 pressure transmitters, seven or eight flow transmitters, and a handful of analytical transmitters. Once flow and other measurements are taken at the pressure stations, they come back through the SCADA system and are written down. We think that automation and AI could help us model and identify useful trends faster and more effectively. We’ll probably also try some generative AI (genAI) later, but those results would be advisory only.”
Get your subscription to Control's tri-weekly newsletter.
Paxson reports that Cobb County-Marietta Water Authority’s IIoT-based, energy-saving projects and data-modeling efforts have been aided by the utility integrating its IT and OT departments into a single technical group about a year and a half ago. They were also greatly supported by upper management‘s approach of bringing everyone into the same room for meetings that are still held monthly.
“IT often had no idea what OT’s job was and vice versa, so it was important to get communications open and get on the same page,” says Paxson. “We also filled one new position for an IT/OT guy, who serves as the bridge point between departments, replaces Ethernet switches, and maintains much of our network infrastructure.”
For other water/wastewater utilities and process users that want to use IIoT and perhaps AI to improve their data models and save energy, Paxson recommends assessing their operations, evaluating existing latencies, deciding how far they want to go, and digging into standards like ISA/IEC 62443 and ISA 100.
“You also have to see how clean your data is, so you can trust what you’re seeing, and make sure your trending is valid,” adds Paxson. “We also proactively and aggressively upgraded our PCs. This is because our critical mission is to provide safe, sustainable and reliable drinking water, so we also have to maintain certain levels of redundancy. Consequently, for us to implement IIoT and possibly AI, we also had to have human backups and controls, who could take recommendations, whether they’re coming from our automation, IIoT or AI, and make sure that what they’re seeing and saying makes sense about what needs adjusting or fixing. While automation technologies bring immense benefits, such as improving safety, efficiency and productivity across industries, these systems fundamentally depend on a diverse range of skilled humans to design, operate, maintain and continually improve them.”
