Transcript
Introduction
Thank you for joining today’s webinar. My name is Matthew. I am a software engineer here at PMI. Today we’ll be covering our white paper on ROI snapshots in Merlin. It’s a new feature that takes you straight to the evidence behind every finding in a power quality report.
If you ever hunted across a recording, clicking through measure after measure and combining capture lists, trying to track down one moment that mattered, this feature gives you that time back.
The Problem: Finding the Needle in the Haystack
Let me start with a scenario you probably already know. You got a recording in front of you. A customer called because their equipment tripped, and the reason is somewhere in this file, but somewhere covers a lot of ground.
Look at everything one recording holds. There’s dozens of measures: voltage, current, power, power factor, unbalance, each across several channels. There are different ways to view every one of them. You have interval graphs and daily profiles and histograms and waveform captures. And there’s even the event list with significant changes in the thousands. Dozens of waveform captures. Event changes, flicker, and/or voltage.
So without Merlin, the hunt looks like this. You pick a measure, pull up its strip chart, you scan it by eye, you look for anything that looks off, then you flip through the waveform captures to see if any of them line up. Then you check another measure to see if that could have tripped the equipment for corroborating details. Cross-reference timestamps across all of it.
You repeat because the thing that tripped the equipment could be hiding in any one of these views. And the whole time, the vast majority of what you’re looking at is perfectly normal. The moment that matters is a needle in this giant haystack of data.
What Is Merlin?
Just a quick bit of context first because some of you live in the Merlin world already and some of you are mostly in PQ Canvas. Merlin is the analysis engine inside PQ Canvas. You point it at a recording and instead of you walking the whole thing by hand, it walks it for you, looking for the phenomena you’d be looking for anyway. Voltage sags, swells, loose neutrals, regulation problems, harmonics, flicker. Then it writes it up as a report.
The Gap That ROI Snapshots Fix
Here’s the gap though, and this is the part that the ROI strip chart snapshots fix. Up until now, Merlin would tell you what it found. The report would say, “Hey, there’s a voltage sag around this time.” Which is useful, but the burden was still on you. You’d read the claim and then you’d go find the proof yourself. Cross-reference the timestamps, pull up the graph, scroll through the window.
The analysis and the evidence lived in two different places, and you were the one shuttling between them. So the story today is really about closing that gap, putting the “show me” right next to the “what happened.”
What Is a Region of Interest?
To get there, I need to introduce one concept because everything else builds on it. It’s the idea of a region of interest. A region of interest or an ROI is just a specific slice of time that Merlin’s analysis has flagged as worth a closer look. The window around a voltage sag, a burst of harmonic distortion, not the whole recording, a tight interval where something actually happened.
And this matters because of that “vast majority looks normal” problem I just mentioned. If ninety-plus percent of the recording is uneventful, the valuable thing isn’t more data, it’s pointing. It’s someone saying, “Out of these two weeks, here, here, and here are the moments that you care about most.” That’s what an ROI is. It’s Merlin drawing a box around the part of the recording that earns your attention.
How Merlin Identifies Regions of Interest
So how does Merlin decide what a region of interest is? The honest answer is it does roughly what a good PQ engineer does. It walks through the recording looking for notable windows, and each phenomenon has its own dedicated analysis. There’s an analysis for flicker, there’s an analysis for sags, one for swells, one for loose neutrals, for regulations, harmonics.
Whichever one of those finds something noteworthy, it records two things: the time range and a short reason why that window is significant. Then Merlin double-checks each window, both numerically and visually, before it stores it.
But here’s the part that I really like because it mirrors how you actually triage. Every finding gets a severity score, zero to ten, which means the regions get ranked. The most critical events rise to the top, so you’re not working through the recording chronologically, event by event, hoping the most important one shows up eventually. You’re seeing the most important events first. That’s the difference between reading a recording front to back and reading it worst first. Worst first is how you want to work if you had the time to score everything yourself.
Snapshots: The Evidence Inline
Now here’s where it comes together. Those reasons are what Merlin uses to build the snapshots, and snapshots are the actual feature we’re here to talk about. A snapshot is a small focused piece of evidence embedded right inline in the report, and it’s not one size fits all. Merlin picks the type of snapshot based on what the finding is.
- If it’s a gradual voltage regulation drift, it shows you a zoomed strip chart because drift is a trend and you need to see it over time.
- If it’s a momentary transient, it shows you a waveform capture because that’s a fast event and you need to see the shape of it cycle by cycle.
- If it’s about a triggered event, it can give you a text artifact, even an event change table that summarizes the cycle-by-cycle changes around the trigger.
And every snapshot carries a caption. That’s the small thing that matters a lot. The caption ties the image back to the finding and says in plain English what you’re looking for and why it’s there. So the events and the reasons always travel together. You’re never handed a graph in isolation and left to guess why it’s relevant.
How to Access Snapshots
How do you actually get to these? Step one, you run a Merlin analysis. This one has already been run, same as always. When it finishes, you can open the overview dashboard, which is what this is. You can select the power quality report here on top. That’s your report.
Now as you read through it, the snapshots are just there inline as clickable links. Everywhere Merlin references a specific piece of data, you click one and it opens in a pop-up right on top of the report. So here’s one that’s an event table. Got a waveform capture. And that pop-up details the whole trait and it opens over the report so you never lose your place in that narrative. You read the claim, click, look at the proof, close it, keep reading. You’re never leaving the report to go find things.
Snapshot 82, waveform-based RMS snapshot of three-phase 3% to 4% voltage step both with weak current correlation. Close it, and I’m right back where I was.
The Snapshot List
There’s one more piece I want you to know about, and it’s near the end of the report. It’s the snapshot list. If we scroll down here, it collects every snapshot from the entire analysis in one place. For each one, it gives you the analysis source, the start and end times, a description, and which sections of the report it appears in. Think of that as your index.
Every entry is clickable. Click one, and it opens that snapshot’s pop-up, same one you’d get from the inline link. So you can go straight to the evidence for the reason you care about. And because each entry also lists the section it appears in, you can know exactly where in the report to read the discussion around it. Instead of reading top to bottom, you can scan the list, find the reason that matters most to your investigation, and pull its evidence up in one click.
Speed and Trust
Let me step back because I don’t want this to sound like just a convenience feature. It’s really two things. The first is speed, and that one’s obvious. You go from “what happened” to “show me” in one click. What used to mean hunting through potentially hours of recording data now takes no more than a click. That timestamp-matching ritual I opened with, gone.
But the second reason is the one I care about more. It’s trust. When Merlin embeds the underlying evidence right next to its conclusion, it’s proving that the conclusion is grounded in real measured data, not just asserted. You’re never asked to take a finding on faith. You can independently confirm it or challenge it using the exact same evidence that Merlin used. And that changes your relationship with the tool. It’s not a black box handing you answers. It’s showing its work, and it’s inviting you to check it.
Summary
So to bring it all home, regions of interest are the specific intervals Merlin decides are worth a closer look. It flags them, it ranks them by severity, and it embeds tight captures of the underlying data with the reasoning right alongside. The result is an analysis where any finding can be verified in a single click without ever searching through the full recording.
If you want to dig deeper into PQ Canvas and Merlin, everything’s at powermonitors.com. Thank you for attending today’s webinar. If you have any questions, you can reach us anytime by calling 800-296-4120 or emailing support@powermonitors.com. Thank you, and have a good rest of your day.
