Solve Costly Out-of-Control Processes with Control Chart Triage in JMP Live (2025-US-30MP-2462)

All right. Thank you for joining this talk on Control Chart Triage in JMP Live. My name is Michael Goff, and I'm a senior software developer on the JMP Live team. And I'm not alone today. Since JMP Live is a collaboration tool, and the feature we're sharing today is a collaboration feature, it'll work better with two of us here. And I have a great co-presenter in Annie.

Thanks, Michael. Hello, everyone. I'm Annie Dudley. I am a statistical developer in JMP, and I'm the developer responsible for control charts in JMP. Today, we are going to talk with you about control chart warnings in JMP Live. We have been working on this feature over many releases, and we've been investing in this effort because we understand that if you're using control charts, you are likely using them to monitor the stability of your process, whether or not it's in control. And when your process is not stable, you often use the control charts to make important decisions in order to correct, take corrective action, bring it back into control. And failure to act on this information can be stressful and costly to your organization.

So our goal today is to help you share information, to communicate and collaborate, and make these decisions more quickly and efficiently. So now, because this is a list of all the features that are now in control chart warnings up in JMP Live, let's go ahead and create a control chart, and then we'll publish it to JMP Live. So here I have some data that I simulated to represent Silicon Carbide Ceramics organization, and we're measuring the flexural strength in megapascals.

So let me start by creating a control chart. I'm going to use the menu, and this is fully supported under Control Chart Builder. So I'm going to use the menu to get into Control Chart Builder, and we're just going to create an individual moving range chart. So I have the flexural strength, and I have the date that I'm going to enter as a subgroup. All right, so here's my control chart. We can see we have some little issues here, but we are already aware of that because we're talking about things that are going on more recently here. But you may notice over here on the right, it says, "Limits were imported from a Column Property." Now, this is going to be really important going forward because the control charts, up and JMP Live, need to have stable limits. They need to not change as a shift is introduced. And so I have set the control limits here as a column property. You can also set them through JSL, through the interface, or through a limits table. So just wanted to mention that. So I'm going to turn off that limit summary. But in order to do the warnings triage, I need to turn on my warnings.

So you can turn on warnings in control chart either by right-clicking in the graph and turning on whatever your favorite test is. But I have more than one favorite test. So I'm going to open the control panel, and I'm going to scroll down here to warnings. And I like test one, which I had already turned on, test two, and test five. And we see that because test two through eight are all runs tests, so those are going to pick up a shift a little bit more quickly than before it goes out of the limits. So I'm going to close that one. But I also want to test to see if anything goes wrong with the dispersion. If we have way too high a variance, that's a problem as well. So I'm going to turn on my test beyond limits. So this looks pretty good. I'm ready to start monitoring my process. So I'm going to click done. And I am now ready to publish to JMP Live. So let's go into the file menu and choose Publish. And first, I need to make sure that my connection is set up correctly. And it looks like I am indeed set up to JMP Live demo.

Okay. And so now let's go ahead and publish. And I tell you people, if I can do this, you can do this. So this is where the magic is happening behind the scenes. So yes, this is what I want to publish. I'm going to publish a new report, and I choose, Next. And we are in Austin, and this is Discovery Summer 2025. So this is defaulting to the right space. And this happens to be a control chart, and we want to look at how triage works. So I'm going to go ahead and publish it in this folder. So let's choose, Next. I'm fine with the title. We could change the title, but that's fine. And so, yes, I want to publish a new data table. So now it's going out and talking to the server. So it looks like we have it up here. So I'm going to click, Open just to make sure that it's there. And sure enough, there it is. So now I'm going to turn it over to Michael so that he can talk with you about what we're looking at here in JMP Live.

Thank you, Annie. Here I am on the JMP Live homepage. And before we talk about what Annie has published, I want to give a quick intro to anyone who's not familiar with JMP Live. So JMP Live is a secure collaboration platform, and it's installed on your own servers at your company. It's used to share JMP content that you create with your colleagues, whether or not they have JMP themselves. Here I am on the default homepage for JMP Live, and we've organized the homepage to help you find the reports that are most relevant to you. We've divided things into a few different sections here, starting with recently viewed, so you can jump back into the reports that you've most recently worked with. We also have a bookmark section for you to save off reports that you frequently visit and want to come back to later. We have a Control Charts Warning section. Now, I'm going to talk a little bit more about this in a few minutes. Finally, we have a recently updated section, so you can stay informed about any new reports that are published to JMP Live. When Annie published her report to JMP Live, she chose a space to publish to.

JMP Live is organized by collaboration spaces. You and your colleagues can decide which spaces you publish to and who has access to what. To get to a space in JMP Live, I have a few different ways to do so. I have a section here on the left called Spaces, and this is organized by spaces that you've favorited with a star, as well as recently visited spaces. I can also get to spaces on the JMP Live app bar, and this is the same list of favorites and recents. And this will be available no matter where you are on JMP Live. So let's go ahead and open up this Discovery Austin space, so we can take a look. So a space is organized by a hierarchy of folders. So I have a couple of folders here in control chart triage in general. If I go to the general folder, we can see a few reports that have been published to this space. Here's the report with the Diamonds data from JMP's sample data library. Let's go ahead and open this up and check it out. When I open up this report, you can see it's a graph builder of price versus carat weight, and we have a local data filter over here on the left-hand side.

Now, you can do more than just looking at a report, but you can also interact with it. I can highlight points if I want to, and I can also configure this local data filter on the left-hand side. For example, if I want to filter on just one color, I can click one color in the local data filter, and the report will update to reflect my changes. Now, I want to talk a little bit about some of the extra benefits you get when you publish your Control Chart Builder reports to JMP Live. And to remind you, this is available for any report based on Control Chart Builder. If you go to the Analyze, Quality and Process menu, any report based on Control Chart Builder, that includes these in the drop-down, from individual moving range all the way down to rare event control charts. These are all supported in control chart triage in JMP Live. So let's go back to the JMP Live homepage. So our goal with control charts in JMP Live is to help you easily find your out-of-control processes. To that end, we have these control charts warning section to outline any reports that have been published to JMP Live that have warnings associated with them.

We mark these reports with a badge to help indicate that there are warnings within. There are different badge colors, which I'll talk a little bit about later. You can also see a list of all reports in an individual space that have control chart warnings. By going to that space, clicking the warnings section on the side rail here, and this presents a flat view of all the reports in the space that contain warnings. Now, when you publish a control chart to JMP Live, and it contains warnings, you can also get notifications about this control chart warnings. For example, if I go to my notifications panel here, I can see that I have a warning about that report that Annie has just published. I also got an email notification to my inbox as well. So these notifications don't get blasted out to everyone who uses JMP Live, but you get them if you're the publisher of the report, you're in a group that a site admin has configured to receive control chart warnings, or you individually opted in to a specific report, a folder of reports, or any control chart with warnings in an entire space. So let's go ahead and open up this report and take a look by clicking the notification link.

You see, we've landed on our control chart report that Annie has published, and just like in JMP, we can see the warnings are highlighted in the report with red circles. I can hover over a warning to see more information about that warning as well. Now, most of what I've shared up until now has been available in previous versions of JMP Live, so now I want to focus in on some of the new features that we've added to JMP Live version 19. To summarize our goals so far, we want to find these out-of-control processes more easily. But in addition to that, we want to learn more about the warnings in our process, and we want to work through these warnings in the process one by one and keep track of our progress while we do so. Finally, we want to stay focused on the processes that urgently need our attention. So let's go ahead and take a look at this report. I want to bring some attention to some of the JMP Live 19 improvements to this report. And one of these is the ability to fit the reports to window in JMP Live. I can go to the snowman menu here, fit to window, and turn this on, and that will allow the report to fill up all of the available screen real estate for us.

So maybe I want this view to persist to everyone, and in addition to just myself. And to do that, I can go to the Script tab here and modify this script to have this fit to window on for everyone. This is the script that powers our report in JMP Live, and I can modify this by adding a line here. Let's just add a line fit to window and turn this on. Add a comma here. When I save this script and go ahead and regenerate the report, it's going to look the same for me, but any future visitors to this report will see the report expanded to fit the window. Other improvements that we've made to Control Chart Builder in JMP Live is the warnings panel. Let's say that I'm a quality manager and I need to know what's going on in this report. What can I learn by looking at it. By default, we've organized all the warnings in the panel into each chart in the report. We have an individual's chart and a moving range chart. I can click the drop-down arrow here to see information about the chart itself. For example, we have one sample out of control, and that represents an alarm rate of 2.5%.

We can also see the warnings that were tested. Like Annie mentioned earlier, she turned on tests one, two, and five. In addition to the summary series of the charts, under each chart, I can see the individual warnings that occurred within them. We can list these warnings out one by one, and that is associated to each warning in the chart itself. The warnings are organized into subgroups here, and I can also expand these warnings and learn a little bit more information about the warnings in the chart. For example, I can see this warning here outside the limits has a measurement of 389.5 megapascals, and one row in the data table makes up this warning in the chart. We can also see which test is failing. In this case, we're beyond the limits, so test one is failing.

We can work through the warnings in this chart in two ways. Like I've done so far, I can click the warnings in the panel to focus in and look at the warnings. I can also look in the chart and look in the warnings this way. If I select a warning in the chart, it will open the warning in the panel as well, so I can jump between them this way.

Now, I have these warnings associated with the report, but what can I do with them? As a quality manager, it's ultimately my responsibility to work with the specific people on the factory floor to make sure this process gets back under control. To that end, in JMP Live 19, we've added the ability to triage your control chart warnings in JMP Live. Let's go ahead and try that out. For this specific warning, we measured a flexural strength below our limits. We want to make sure this has been dealt with. To do that, I can right-click the warning and click Triage to open up the warning in the panel. We can see this triage section underneath the details of this warning, and a triage is made up of three fields. The first field is the status. We start in the open status, and this is what any new warnings that are generated will start in. This means that something has gone wrong, and no one has figured out what has gone wrong with it yet. We also have a few other statuses to choose from. We have investigating, meaning that someone has identified this problem and is looking into why it's going wrong.

And finally, addressed, meaning that someone has figured out what went wrong, and we've taken corrective action to address the issue. We also have an Assignee field, and I can quickly assign this to myself with one click, but I'm not on the factory floor, so I don't think I'm the best equipped to handle the problem that's going on here. So I'm going to assign this over to Annie to take a look. Now, I happen to know that Annie has JMP desktop on her machine, but that doesn't matter on my decision to assign this warning to her. I can assign this to anyone who can see this report in JMP Live, and they don't need to have JMP on their computer to deal with it. So what's going on with this warning? We have a notes field, so I'm going to leave a note to Annie to be polite. "Annie, something has gone wrong in this process. Could you check it out?" All right, everything looks good here, so I'm going to go ahead and save this triage to the warning. Down in the moving range chart as well, I can see this happened at a similar time frame.

So I'm going to go ahead and triage this one as well and assign that over to Annie. Let's go ahead and save that. So now that Annie has been assigned these warnings, let's check in on her progress and see what's going wrong.

Thanks, Michael. So I just received an email notification from Michael that I was just assigned two warnings. I can also see up in the notification section that Michael just assigned, just a minute ago, assigned two warnings to me. I can also see over here in the warnings panel that these two signals are assigned to me. But what happens if I ignore all that? What happens if I'm not looking at my email? I don't think to look at this warnings bill. Maybe I have lots of things going on at once.

Annie? You need to share your screen.

Oh, it says I'm sharing it.

I'm seeing it.

Oh, I'm not seeing it.

Okay, I'm going to stop and I'll restart. Okay.

I see it now.

Okay, I'll start over. Okay, so I just received a notification email from Michael. I also can see just a minute ago, Michael assigned these two warnings to me. I can see in the warnings panel that both of these warnings are now assigned to me. But what happens if I ignore my... I'm not looking at my email. I ignore this bell. Maybe there's other notifications coming in, and I ignore this because it doesn't occur to me that it applies to me. What happens if I miss all that? So we've also added a new warnings list. So if I go to the home page and I click on this new warnings list, this comes in really handy because they make this into a to-do list. So everything that I'm assigned shows up in this list. And so you can see here, this is the assignee is me. And anything that's under investigation or is currently open, shows up in this list. And this is across all charts. So this is really handy. So I'm going to go ahead and go to the first item in my to-do list, and it takes me right back to the chart. And in fact, it opens that particular warning. So that's really handy.

So now I happen to know what's going on here. So I'm going to replace these notes. His notes will be saved, but I'm going to add to that and say that this was a silicon sand mixing issue, and it's been fixed. That's why we don't see any more problems later on down. And I'm going to change this to addressed. And lastly, I need to hit save. So you can see it's turned green. It's also turned green over here, which is a nice, happy thing. Now, let's see. I have one more. So you may recall that we're looking here at the moving range. And so this moving range is calculated, because we have a range span of two, as the distance between this point and the previous point. And that's a pretty big distance right there. Granted, this point is big, too. But that guy goes above the center line some, and so that's a larger distance. And so that's why. And so it's all associated with this same sand, silicon mixing issue. So I'm going to say that that was the same. "This was also the sand, silicon mixing issue, which has been fixed," and I'm going to correct the spelling on mixing. Okay. And I'm ready to hit save.

Okay. So now I have addressed both of the issues that I was assigned. I can see they're both green here in the chart. And now I can see the warnings for the overall status of the chart has turned green. And so now you can see we have an effective way of communicating with our colleagues, all of this information. For each point, we can see the status that it's been addressed, who addressed it, who investigated it, and any notes associated with this. And none of this information is going to be lost in a hallway conversation, or on a sticky note, or even in an email that might get deleted after a few months. This information is there on the server associated with every report, and anybody who has access to view this report can go back and see all this information. And so now I'm going to turn it back over to Michael, and he's going to show you more details about what exactly is being saved.

Thank you, Annie. I'm still on the report page here, and I can see that Annie has addressed all the warnings in this report. I can see those green squares in the chart, and I can see the address status on the warnings over here in the panel. If I wanted to further reassure myself, I can use the filters in the warning panel and exclude all of the addressed warnings from the panel. When I do that, I can see that there are no more warnings left in the panel. Let's go ahead and clear this filter. Let's say I wanted to figure out what was going on with this particular warning over time. To do that, I can open the history panel here and view a history of everything that's happened with this warning over a period of time. We can see who assigned this warning and what they said, who addressed this warning and what they said. Now, this example has been pretty straightforward, but if there was additional back and forth figuring out what went wrong with this warning, I can see that in this panel as well. Perhaps most importantly, I can see when everything took place for the purposes of auditing.

Now, this history is available in the JMP Live user interface, but I can also choose to download a CSV file containing all this information as well. And since it's a CSV file, you can open it in really any application that supports CSV. But I'm going to go ahead and open this up using JMP to take a look. So in JMP, you can go to File, Open, and download this CSV file. So this download is organized into two different sections here. The first section is a whole bunch of attributes about this warning to help you uniquely identify it within the report. And then underneath is a history section. This lets us know everything that's happened with this warning over time. We can see that the warning was created. We can see it was then updated by myself, where I assigned it to Annie, and finally, where Annie updated it and addressed this warning. So that's the history of one warning, but I can also choose a different download in the JMP Live interface. If I collapse this warning, up here at the top of the warnings panel, I can download the current state of all warnings in the report as well.

When I click this button, I have some options to filter the warnings in this list, but I'm going to take the defaults for now and go ahead and prepare this download. Just like before, I'm going to choose I'm going to open this in JMP, but since it's a CSV file, you can open it in any of your preferred applications. You don't need JMP to open this. In JMP, I'm just going to go ahead and open this one up as well. In this CSV file, we have one line representing each warning in the list. We have a triage status. We can see that both of these warnings are addressed, and we can see that Annie is the one who addressed them, as well as the notes that she wrote down about this warning. If I scroll over to the right, I can see all those uniquely identifying attributes about the warning as well, in case you need to identify it within the report. Let's go ahead and close this download and go back over to JMP Live. Overall, we can see that this Silicon Carbide process is going well. Are there other processes on JMP Live that I need to worry about?

To check that out, I can go back to the home page and go to this warnings list. Now, Annie already alluded to this, but the warnings list represents a to-do list, but that's just the default state of this warnings list. I can choose to change the assignee if I wanted to look at someone else's warnings. So I can, for example, go down and check out Annie's warnings by typing in Annie. And we can see that she's already addressed everything that she needed to take care of. And I can confirm that by changing the status over here to include address warnings as well. So I can see those two warnings that she just addressed. I can also see an overall to-do list. So let's go ahead and get rid of address, so we don't overwhelm ourselves. But I can also clear out the assignee entirely and see all of the warnings on the entire JMP Live site, regardless of the assignee. I can see some of the warnings are assigned to myself, but I can also see that some warnings are unassigned. I can also filter on unassigned as well. If I want to find warnings that need to be taken care of, I can find the unassigned ones and jump in and assign them to someone.

Now, this page is dedicated to specific warnings, but what if we wanted to look at warnings from a higher level? We can do that by going back to the Overview page and using this Control Chart warnings section. I can click, View All here to view all content on JMP Live and have it be prefiltered by control charts that have warnings. The report we've been working on has been green, and that means that Annie has moved all these reports to addressed. But what about the other statuses here? Now that we're tracking individual warnings in a report, we can now determine the overhaul triage progress of a report as well. We have the green one here that we've worked on and fully addressed, but we have two other colors as well. We have yellow here that represents warnings under investigation. This means that there are no more open warnings contained within this report, but there's at least one warning left that is being investigated. Then, finally, we have the red badge, and that means that this report contains warnings that are open. Now that we track the overall triage progress of a report, I can choose to refine these filters as well.

Maybe I want to look at only the reports that are being currently investigated, and I can do that and filter out all the others. Let's go ahead and add that back. And I can also use a filter of warnings assigned to me to track reports where I have work to do. Let's go ahead and get rid of the assign to me filter. So what's the benefit of the triage progress here? By tracking the triage progress, we're able to give you fewer notifications on JMP Live. We can stop sending out notifications if your report no longer needs your attention. Let's say you have a report on JMP Live, and JMP Live has already notified you about this report. When the report is updated, JMP Live will reassess the triage progress of the report. If it's still under investigation or fully addressed, we're going to assume that you're on top of things, and we don't send you additional notifications. However, if you had a report that was fully addressed and then comes back to needs attention, we're going to send out new notifications about these new warnings. While I was talking here, we can see that the Silicon Carbide report has been updated, and it went from green to red, and that means that we have new open warnings.

We can see in my notifications panel that I've got new notifications about this report as well. Let's go ahead and open up the report and take a look. It looks like this report has changed, and we have two new warnings at the end of the chart. It looks like these are failures of a runs test, so we're not quite outside our limits yet, but we might soon if we don't address the issues. Now, why did this report update? To see why, I can go over to the data tab, and this is a section of all the data tables that power this report. In this case, the Silicon Carbide Ceramics data table has updated one minute ago. Now, when data is updated on JMP Live, either manually or maybe it was set up to use a scheduled data refresh where JMP Live can pull data from a database or another server on a schedule, either way, when this data updates, it will then update the report that uses this data. That's where these new warnings came from. Now, these warnings both happened on October seventh and eighth, so it's in the last day or so, so perhaps this is the same issue.

I can triage these together. To do that, I can draw a box around these warnings. When I right-click and triage, I can triage both of these warnings at the same time. Maybe Annie has gone home for the day, and someone else is on call on the factory floor, and I need to assign this to someone, whoever's on call. I don't really want to assign this to an individual, but instead, I want to assign it to a group of people. To do that, I can just type in a group. We have our Ceramics Process Engineers Group, which comprises of everyone who's on call. When I assign this to this group, it will notify everyone in this group that the warnings have been assigned to them. When I click Save, I can jump over to the warnings panel and see that both of these new warnings have been assigned to that Ceramics Process Engineers Group. So let's go ahead and check in on one of those engineers as they investigate what's going wrong.

Thank you, Michael. So it turns out I'm still here. I have not left for the day. And I can see that I got another email, and I can see in my notifications that Michael assigned these warnings to the group that I am a member of, the Ceramics Process Engineers. So now I can see that these two are showing up as a shift has occurred. Okay, we can see that shift right here. Now, I just wanted to remind everybody, remember, I had fixed limits, static limits that I had set in the control and the column properties at the beginning. And this is the reason for it, because if a shift happens, and we're going either above the center line or the shift is going like, say, one Sigma below the center line, then that can cause the overall Sigma for the chart to increase, which would then increase the limits, which would mean this point, which we have already triaged, might not be out of control anymore. So it's going to cause all the prior triage information to get all wonky. Stuff could go in control or out of control. So if shifts happen, we don't want to mess with prior triage points.

Okay, so that said, I am here, I can take a look at this. I can highlight this the same way Michael did and choose Triage. Now, I don't actually know what's happening. I need to investigate this. So I can change it to investigating, and I can assign them both to myself so that everybody knows they don't have to drop everything and go investigate. I'm on it. Okay. So today, we have shown you some new features in JMP Live that we think will help you with your processes. It'll help you find out-of-control processes and signals more quickly and easily. It'll help you learn more about what's going on with each and every warning because all the notes are there. It'll help you work through each warning one by one or in a group, like we showed you here. And for each warning, it'll help you communicate to anybody who has access to seeing this chart, what the current status is, who the assignee is, who was responsible for investigating this, or is currently responsible for investigating it, and any notes that everybody has found along the way as investigations have been proceeding. So what do you need to do?

All of these new abilities here, are here if you want them. You don't have to use them. You don't have to do anything. If you want the basic benefits, though, of control chart support in JMP Live, publish Control Chart Builder reports up to JMP Live. There are some pretty nice options here in this list that we've been building over the last five releases. These notifications where you can opt out if you don't want to receive information about a chart, or you can choose to, for example, opt in if you want to pay attention to a particular process that's going on. We now also support a separate limits file so that you can easily manage your limits there. But if you want the warnings triage, publish your Control Chart Builder reports to JMP Live that have static limits. Again, you can set those static limits through a limits file, through a column property, or through JSL. So in this list of benefits that we've been building up over time for control chart, support, and JMP Live, we have this one little line here that we think represents a pretty big step forward in helping you understand what's going on in your processes and helping you collaborate more effectively with your colleagues all across your organization to solve issues and get back to the business of running your business.

Now, we did not do this on our own. Michael and I were joined by Aurora, Justin, and Praveena. And we'd like to thank product management for guiding us along the way. And we need to thank the testers who banged on the product pretty rigorously and the JMP employees and customers who've tried it out and given us feedback along the way. We'd like to invite you to please stay for questions and answers, and come and talk to us wherever you see us at the conference. I'd like to thank you very much for your time.