Secure Connectivity Brings Onsite Insights to Offsite Services

Juan-Ramon CamposThe J. J. Pickle Research Campus is a collaboration between the University of Texas (UT) and industry, including Emerson Automation Solutions. The pilot plant was the first location to implement Emerson wireless, and its current focus is separations research, including the removal of carbon dioxide (CO2) from stack gas. The plant also uses Plantweb Optics to monitor plant controls on distillation, CO2 adsorption/stripping, liquid-liquid extraction, oil extraction from algae, and verification of separation.

The facility is also used for process scale-up and troubleshooting, and as a beta site for much of Emerson’s new technology. It has been running DeltaV distributed control system for 18 years, with 600 DSTs running several units, an M-series controller and I/O, S-series controllers and I/O, Electronic Marshalling, AMS Device Manager and machinery health monitoring, and more than 60 wireless HART devices, plus wired HART and Foundation fieldbus devices, monitored by AMS Device View.

The plant runs with two electricians, one mechanic and a chemist. “Before Plantweb Optics, we’d come in, drink a lot of coffee and open up AMS Alert Monitor, DeltaV Explorer, Wireless Gateway Web Interface, Control Studio and DeltaV Operate to determine the health of the plant,” said Juan-Ramon Campos, instrument technician, electrician and operator. “If abnormal conditions existed, we’d use AMS, Valve Link, Insight, Control Studio and Process History to troubleshoot and determine the next course of action. After determining the problem, we’d head to field with a 475 [field communicator], multimeter, service manual and tools, and after correcting the problem, we’d return to the control room and use above mentioned programs to verify performance.”

Campos and co-presenters Tinh Phan, digital transformation solution architect, Emerson Automation Solutions, and Dhawal Tyagi, chief product officer, ioTium, spoke at the Emerson Global Users Exchange this week in San Antonio.

Remote support couldn’t stay away

“I was having to drive to J.J. Pickle for asset management, which was involving too many long days on the road,” said Phan. Emerson also needed to secure remote access into J.J. Pickle for diagnostics of DeltaV systems and wireless systems.

“We also wanted to do it for implementing new Emerson technologies. I couldn’t see remotely how the instruments were working and how information was being passed down,” Phan said. The project goals became:

  • Get real-time remote access for a view of plant reliability and operational performance.
  • Integrate multiple technologies for a holistic view of asset health and to support recommendations.
  • Have a secure, yet flexible architecture to interface and integrate with other platforms.

Security at the edge

“People need data for analysis from a lot of different pieces of equipment with different protocols,” said Tyagi. On cybersecurity, “The hackers are always two steps ahead, and nobody has a perfect system. You need layers, like you have at home—a lock on the door, and the valuables in a safe.”

Tyagi described two ways to deploy edge applications—on a cloud, or in the factory. “If you do it in the factory, you still have to get the data to the applications, so how do you connect it to the cloud? And if you try to keep it at the edge—in the factory—you need a Linux or Windows machine, you have to install it and maintain it, and you may need hundreds of them for different locations.”

Instead, to add a monitoring or control application, ioTium installs “iNodes” at the plant, and an “orchestrator” that can access the nodes. Nodes can be connected to private clouds at the company, third-party analytics, instrument vendor remote facilities, etc.

“The resulting private network is not part of the internet,” Tyagi said. “Each iNode is in isolation—any hack can only access one node, one network. You can minimize risk because you know the boundary of the attack.”

End run around IT

Emerson performed a proof of concept using iNodes in iOps, and tested it in Austin before deploying it at JJ Pickle. “We used Azure, ran an executable on AMS Device View, and immediately had a view,” Phan said. “We used a virtual iNode at J.J. Pickle and a hardware node at Emerson. We were able to avoid opening too many ports at J.J. Pickle, which UT IT would not have wanted.”

By securely connecting remotely to the control systems, Emerson is able to monitor all the smart instrumentation. “Before, they were fighting fires every day at the plant,” Phan said. They could commission many devices on a remote monitor and drill down through AMS Device View to see problems and calibration status. “But they still could not do everything on premises, and there was too much to do.

“Plantweb Optics showed a humungous amount of compromised devices that they can now fix—many low batteries, devices not running, etc. that they can repair or disable.”

From JJ Pickle’s point of view, “Installation went seamlessly,” Campos said. “Now we have real-time capture of data without disrupting critical assets. I can access it from my house, see status, make a plan, and get faster resolution of device problems from Emerson.”