In the process of covering automation and the related production industries, we get to see a lot very interesting new technologies. Some of it is jaw dropping, but not very practical. Others are practical but not very inspiring. In the case of Beet Analytics’ Envision predictive maintenance (PM) software, I’ve seen something that is both intriguing in its approach to a problem and practical in its application.
You may not yet have heard about Beet Analytics. I met with them for the first time just last year at the Profinet Executive Forum (read my initial report about Envision while it was still in beta testing). What interests me most in my meetings with David Wang, CTO at Beet, is the company’s approach to the PM issue.
“In a typical control system, sensors are connected to the PLC to determine if a condition is met,” Wang explains. “If the condition is met, the PLC goes to next step.” Beet’s Envision application “looks not just at if a condition is met, but how much time the process to meet it took.”
Wang says that by examining this aspect—the time it takes to meet a condition in the motion and control realm—you can see the heartbeat of automation. To make his heartbeat analogy, Wang connects the graph of time stamp data created by Envision to an EKG graph of a heartbeat.
“Just like we use an EKG to diagnose heart problems because it’s a fast and simple-to-use technology,” says Wang, “Envision focuses on the heartbeat of automation—motion and control—to identify problems before they occur in a fast and simple way.”
Wang explains that the difference between Envision and most factory information systems is that Envision looks at every step between sensor and controller and notes when actions begin to drift out of spec; this allows users to identify the exact moment when drift begins. Users can then drill down into that specific point to see what’s going on in granular detail.
“Factory information systems look at the overall machine cycle—not the individual time stamps that comprise the cycle,” says Wang. “The difference between such systems and Envision is the granularity Envision provides to see precise variations in system actions at the earliest moment when things start behaving out of spec so that corrective action can be taken earlier to prevent downtime.”
To deliver this level of information, Wang says Envision captures not just motion-related data, but any handshake between devices or systems. To capture this data, Beet inputs code into PLCs to capture time stamp information. Wang says this code does not impact the logic or motion functions of the PLC.
“It's a separate function block to time event triggers and stops,” he adds, referring to the code Beet inputs into PLCs to interact with Envision. “With this approach, Envision can capture any and all performance variances.”
Wang says that Beet typically installs the function block code. However, the company also offers classes to teach users or third-party integrators how to install the function blocks themselves.
Capturing data through code is another big difference between Envision and other PM tools. Wang says that most other PM tools use “special sensors to detect vibrations or variation to deliver their predictions. This is a costly approach because additional sensors have to be added the system and they are not always as accurate they should be.”
With its code-based approach, Beet adds no physical hardware to existing systems to capture data.
The video below illustrates how Envision works.
