Industrial edge devices begin to show their value
Is there any truth to the hype around edge computing? As a reminder, that term refers to a distributed computing concept where data processing and connectivity are available at the local process level rather than squirreled away in the network core or in the cloud.
For a machine builder, the idea might seem trivial, since equipment often includes embedded control, communication, and memory functions. Therefore, it’s important to understand that the key word in the concept is distributed, meaning that edge computing devices are capable of working together with other devices and systems. Thus, edge computing is strongly related to another much-hyped term: IoT, the internet of things.
Does that add up to a mountain of fluff, or is there a natural synergy between the two concepts that yields real value for equipment designers? Although both concepts are relatively new, the following real-world examples demonstrate how they work hand in hand and the potential value they can produce.
Engenuity, Inc. (engenuityinc.com) provides control automation and data integration solutions across several industries, with a primary focus on advanced technologies for Oil & Gas. Through its work in this industry the company design team identified specific opportunities around validation pressure testing of blowout preventers (BOPs) and well control equipment.
Pressure tests must be completed every few weeks in order to maintain the safety and integrity of drilling facilities. They are typically executed and recorded manually, taking hours to complete. Since it can cost as much as $6.00 per second to operate the associated valve arrays in an offshore drilling application, testing can cost millions of dollars annually.
In collaboration with customers like Shell International Exploration and Production Company, Engenuity developed a set of fully automated test execution and reporting products. Engenuity’s EZ Chart, EZ Valve, and EZ Vision work together to precisely pressurize equipment, capture critical process parameters, validate the results against predetermined criteria, actuate up to 70 valves through as many as 30 test sequences, and pinpoint leaks in the system using innovative acoustic detection methods.
Depending on an installation’s size and operations, Engenuity’s solution can save users 10-20 hours for each full test with a fully integrated system. Their high-reliability solution uses an edge programmable industrial controller (EPIC) for local control and data processing, storage, and replication. Data collected from these systems is stored in a MariaDB instance (an open-source database) running on the edge controller and then replicated to a twin controller to ensure 100% uptime and easy data access. Collected data can be mapped between tests for comparison and reliability purposes and used to generate PDF reports.
According to Jeff Hilpert, president at Engenuity, “the integration of data and control on a single backplane [using an industrial edge controller] is key to dramatically lowering cost, improving uptime, accessing data, and expanding utilization.” The approach allows for continuous validation in the event of network or HMI failure, ensuring that no data loss or process interruptions occur during testing.
Shumaker Industries, an OEM that manufactures automated truck wash systems, wanted to add a remote monitoring option for customers who were interested in a rent-to-own arrangement. However, the programmable logic controller (PLC) they used to provide control didn’t have the ability to perform the database transactions they would need to track usage. Martins Electrical Service (MES), a Pennsylvania-based system integrator, was hired to design this new equipment-as-a-service feature.
Initially, MES experimented with a proprietary remote monitoring system—wireless sensors that pumped data into a cloud database through a local IoT gateway—but the system could deliver updates only every 10 minutes, and that data was locked in the vendor’s private storage system. Then MES heard about new options in industrial edge devices and decided to try another approach.
Rather than installing additional sensors, MES used an edge I/O module to capture existing sensor data in parallel with the wash system’s PLC. Using the IoT platform embedded in the module—IBM’s open-source Node-RED—they organized that I/O data into timestamped sets associated with a vehicle number. Then they connected a 4G cellular router to the I/O module and pushed the data structures directly into cloud-hosted storage, which they had rented from MongoDB.
This arrangement allowed them to deliver a scalable monitoring system without worrying about vendor lock-in or IT infrastructure maintenance. And by simply grabbing existing I/O signals, they avoided the need to communicate with or modify the manufacturer’s existing PLC controls.
To complete their EaaS design, MES built a web-based dashboard that let Shumaker visualize the wash system in operation with one-second resolution. The system also generates email and text alerts in Node-RED, including a monthly report on the number of trucks processed by each system, which Shumaker uses to calculate billing.
“[Edge I/O] is perfect being that it…has all the security built in,” reflects Leslie Martin, automation technician at MES. “It’s a small add-on, so it’s easy to make an option for customers.”
From one-size to enterprise
In the early months of the COVID-19 pandemic, in response to increasing demand for hand sanitizer, Northeast Automation Company Inc. (NACI; northeastautomationco.com) was hired by a new firm called Emerald 66 to help them set up an automated bottling and packaging process. Knowing that they were competing against low-paid, high-volume workforces, their goal was to use technology to create a competitive advantage. NACI decided to use a combination of edge devices to capitalize on flexible integration options. Within three months NACI had 15 pieces of equipment running and reporting process data to a central database.
Initially, Emerald 66 had built their business around distribution for a single large purchaser, so throughput was the critical process metric. But the situation changed significantly when Emerald 66’s customer suffered a financial setback and had to close production. Emerald 66 pivoted their whole operation to become a multi-product facility. Automation grew from processing a high volume of one-gallon, single-formulation containers to working with a variety of sanitizer chemistries in different batch sizes and packaging form factors: from small two-, four-, six-, and eight-ounce containers, hand pumps, and spray bottles, to large jugs in excess of one gallon.
They also found a valuable niche in original equipment development for overseas export. In one example, using an edge controller, NACI took an inline mixer design from concept to implementation in about five days, including a mobile operator interface they built using the controller’s embedded HMI server. By their estimate, they implemented the controls in about four hours for a system Emerald 66 could sell for $50,000.
“The fluidity and dynamics of modern manufacturing requires extremely fast response to changing market demands…” says Tom Coombs, principal engineer at NACI. “[Edge computing] puts dynamic manufacturing data at the edge of the production line and into enterprise systems simultaneously in real-time.”
Doing more with less
Given the economic pressures and increased pace of tech adoption that continue as the world grapples with the global pandemic, there are many reasons why equipment designers are giving a closer look to rising technologies like edge computing.
With the traditional automation technology stack, Engenuity’s design would likely require a networked industrial PC (IPC) to store and replicate process data, requiring additional configuration and imposing operating system (OS) licensing and maintenance costs. By opting for an edge controller, they simplified their design, decreasing costs and increasing reliability.
MES also used edge computing to reduce complexity. With a single edge I/O module, they were able to process I/O data, open a secure connection to cloud services, execute database transactions, and generate reports. The final design opens up a new business model for their customer with equipment that is cost-effective to deploy and maintain.
Similarly, NACI’s example demonstrates the rapid pace of development that industrial edge devices support. Because each component in their system was capable of handling sensing, control, data processing, and database transactions, the process could be expanded and reconfigured dynamically to keep up with the pace of change. They were also able to quickly roll their technical expertise into creating additional revenue streams.
With embedded general-purpose computing, networking, and storage, industrial edge devices can assimilate many automation functions that traditionally required multiple devices and more complex configuration. Edge computing invigorates traditional equipment design by providing the connectivity and data processing necessary to participate in a network of intelligent devices, without the overhead typically required.
Filed Under: IoT • IIoT • internet of things • Industry 4.0