Hands-On with AI: Bringing Senseye to the Classroom- A Panel Discussion
Show notes
Welcome to the Trend Detection podcast, brought to you by Senseye Predictive Maintenance – which gives you visibility and insights into all your assets, from single machines to full plants to help you reduce downtime, increase knowledge sharing and accelerate digital transformation across your organization.
We are joined by Dr Aris Alexoulis from Manchester Metropolitan University and Jordan Walters and Steve Jones from Siemens to discuss:
How Siemens’ Connected Curriculum equips students with real-world Industry 4.0 skills by integrating Senseye Predictive Maintenance into university teaching.
The structure of the three-week pilot at MMU: from ingesting data to configuring anomalies, trends, and thresholds, giving students hands-on PdM experience.
The role of Senseye Copilot in supporting learning—using natural language to explain alerts, provide context, and extend teaching beyond the classroom.
The impact on students’ employability and career pathways, with predictive maintenance skills shaping projects and future opportunities.
How academic feedback drives product improvements—such as simplifying derived measures—and helps refine training for both students and industry users.
Next steps for the program: embedding the module at MMU, scaling to other universities, and building a distributed network of Festo labs for collaborative PdM monitoring.
You can find out more about how Senseye Predictive Maintenance can reduce unplanned downtime and contribute towards improved sustainability within your manufacturing plants, by visiting: www.siemens.com/senseye-predictive-maintenance
Transcript
Pretty much this was all covered by Steve, but what I wanted to add is that there are different perspectives. There's industrial perspective of the tool. There's the academic perspective of the technology, not necessarily the tool. And then mining the two. I think that's where connected curriculum, um, is at its strongest, because again, I'm trying, trying to kind of solve both sides effectively, uh, the perspective of the other, kind of blend that into a useful experience for the student. Welcome to the trend detection podcast powered by Siemens, and now your host Nel Sullivan. Hello and welcome to the latest episode of the Trend detection podcast. If you're watching the video, obviously the audio won't be able to see this, but if you're watching the video, you can see we're in a very cool location today. We're at MMU um and we're actually following on from a previous podcast we did with Steve just here about really exciting initiative and connected curriculum. Um, but I guess before we go into the topic, I'll just ask each of the panelists today to introduce themselves. So if I start with Steve, and just your role, and, and maybe a little bit about connected curriculum as well, just to kick things off. Yeah, no problem. So my name is Steve Jones, I'm the connected curriculum lead. Uh, we've been doing this for 6 years now, even the Connected Curriculum Initiative, and we have 21 partners, and basically it's aimed at helping our academic partners understand Siemen's view of the world of Industry 4.0. Fantastic. And Iris, uh, my name is Isa Leoulis. I'm a senior lecturer at Manchester Metropolitan University. I'm a program leader of a program and a master's program called Eening SM Systems, and another called Robotics Information. Both align well with connected curriculum and have been the link for the relationship with Siemens for a few years now. Fantastic. Last but not least, Jordan, uh, I'm Jordan Walters. I'm with, uh, Siemens Sensai, uh, and I work on the delivery for Europe Plus, and, uh, I've had the good opportunity here with Steve and Aris to, uh, work on the connected curriculum project, uh, bringing Sensai to students. Fantastic. Well, thank you all for being here. We've got nice sort of diverse range of views and, um, yeah, and background on this, so really fantastic. Look forward to the discussion. Um, if we go straight into it, I guess Steve, we could pick up from our sort of last conversation because we were describing what connected curriculum is and the use of Sensi. We started to get into that, but, um, obviously since then there's been initial sessions with students. So how, how have they gone? How have they been received? Yeah, it went really well. So we did it here at MMU with, uh, we did it over 3 weeks, I think it was, with Harris and the support of Jordan too and other members of the connected curriculum team. Um, went really well. We had a group of students who followed along with that project and implemented it, and the whole point of that is to give them practical exposure, real hands-on exposure to the tool. So yeah, with Iris's backing and, and, and the input from Jordan and the team, I, I think it went really well. Students successfully managed to do the activity over those weeks, and, uh, I think they gained a lot from it. Fantastic, and that, that, I guess there's some personal interest to me actually, you know, in terms of picking up. Sense of the tool, I mean, what was the, the personal feedback about the tool itself? Yeah, it was good, yeah, I think the students enjoyed it and you know, ultimately they, they we were successful with what they did and, and often the proof's in the pudding with these things, isn't it, you know, it's. So they built it over a number of weeks and used the tool and, and set up the rules, etc. um, and then interrogated it at the end when uh with the different um attention levels that were raised. Um, and it, it went really well, you know, the capable students, you know, the mature MSC students, um, and they did really well, and obviously the support of the team in making what is quite a complex topic understandable and achievable, was, was really great, and to see the fruits of that with the students being successful with what they did, was, was, was really great to see. No, fantastic, that's really great, and. And Aris, from your point of view, um, what's been the reaction from the students and your conversations you've been having with your students? So the students received it really well. Uh, I think in the beginning they were a bit, um, they were a little conscious of the complexity of the technologies involved, but I think going in, I really appreciated. Actually, how straightforward the team made it look like, although this is really complicated stuff, but it was presented in a way that was easy to digest. It was well, um, the equipment was well designed and split, so to provide both hands on training but also the background information in a um succinct enough way. And I think coming out of the series of sessions, they all appreciated having the experience with an industrial grade tool that does, you know, that that that's been deployed in industry as we speak. And been trained by people who actually work in the industry, having businesses. So again, having that skill set, they kind of all understood and appreciated that. I guess, um, and maybe Jordan, you could answer this in terms of how, you know, because you get what you're saying it's about the complexities about, you know, learning the software and how it applies in that industrial environment. So how did you sort of go about that, Jordan, so I guess you were heavily involved on that side, maybe you could explain your role that first of all. Yeah, for sure, yeah, yeah, I mean, that's a really good question. I think it, it builds on both from what Iris was saying, right, the um, the application of Sensei is typically in industry, so we're operating on live data. From specific machines that, you know, are well understood, have continuous data typically, um, and allow us to kind of work with the experts for that machine to identify the key condition indicators that we need to monitor in order to detect their failure modes. So, obviously, pivoting this over to students who, you know, aren't machine experts, uh, to potentially data that isn't continuous or is a lot more, uh. Synthetic is a bit of a challenge. So what I did in this case is I generated some data sets which had different kind of failure modes. We tried to keep it pretty basic, uh, you know, the three, I'd say most basic detections we have in the anomalies, the trends, and to some extent, thresholds. Um, and this just meant that it kept all of the configuration quite low in terms of what was required, so it keeps it nice and simple because. You know, as Iris was saying, there's lots of er parts of Sensai which is gonna be. Quite tricky to set up if you're not a machine expert or even, you know, if you haven't had that much experience with Sensi like the right measures for example. And I guess, um, as well, so you all have a very important role in this. So how did you all sort of collaborate together to build up, build up the program, and obviously execute it in the, in the classroom environment we're, we're in today, me and Steve, and stuff. Yeah, I mean, obviously it's a huge challenge because not only have we got the, the technical complexities of using industrial tools that, you know, uh, we are still learning from a, from an academic point of view on the, on the connected curriculum side. You've also then got the educational aspect of that, you know, you need to, to be able to successfully. Uh, you know, deliver something educationally, you need to really understand the technical aspects of that, and, and that's why having Jordan as part of it, made it just so much better, cos Jordan's got obviously a, a real deep and rich understanding technically of what's going on. And then having Aris as a as a current practitioner and a lecturer at MMU who've got the real sort of ability and skills in explaining those um those topics and those technologies to students. So it was, it was a great mix of of bringing together different skill sets in this environment, the technical. And the educational expertise to be able to produce it, so, um, but, you know, it's a huge challenge and, and it's really hard for learners to get experience of these tool sets and something as cutting edge as Sensi in any other way, you know, it's quite, it's quite complex and you know, for academics to get access to these tools is quite unusual really, so. You know, really proud of what the team did, and I think the fact that it was successful and the students enjoyed it, and they were successful with what they did, just shows what a great job the team did. Yeah, and maybe from your perspective as well, Harris, it'd be great to you. Yeah, absolutely. So I mean, pretty much this was all covered by Steve, but what I wanted to add is that there are different perspectives. There's industrial perspective of the tool. There's the academic perspective of the technology, not necessarily the tool, and then mining the two, I think that's where connected curriculum is at its strongest. Because again, I'm trying, trying to kind of solve both sides effectively, um, the the perspective of the other and kind of blend that into a useful experience for the student and again also. Adjust the level of complexity of the um top material to the respective audience. Uh again, yes, it was master's students, but they were not master students in predictive measurements or massive learning. So again. Understand the characteristics of of the the the the current knowledge or level of the students on a specific topic and just building a research around that. It's quite important from the ground up but also without. whilst keeping them engaged, not overcomplicating things. So it's a fine line um to cover. Yeah, because I guess there's a, there's a danger. Maybe, maybe you know, disengagement if you're introduced at all, and it's, you know, without, you know, any, any background or any, you know, time taken to actually teach them about it and how it works, things like that, it can overwhelm, I guess is the thing, no matter how good the technology might be and if it's AI or whatever. And again, from a, from a technical process perspective, this is the bread and butter. Sometimes taking a step back from the technology and what that means is not always something that you know, um, colleagues have to do for the day to day sort of work. So again, having that perspective for the engineers and then obviously for the academics that it's not only about the theory or the tools of mathematical models behind it, it's also the application, the industrial relevance of the tool. And again my theory is quite, quite important. So I guess that's a bit, the really big policy, you know, telling AI theory in into practice, right, I guess that that's the opportunity you're you're giving or you're all providing here. Yeah, absolutely, and again, having access to tools at the cutting edge of technology that would not necessarily be seen in academic environments due to various restrictions. So again, having that experience, um, it's quite important. And in terms of, I guess sort of think about direct feedback from, from students. I know we talked about, you know, you know, generally, you know, really, really engaged and things, but did they, you know, when they first started using Sensei, I mean it might be a question for yourself, Jordan, you know, introducing the tool, but what sort of feedback did you get directly from students about Sensei itself? Yeah, well, from, from my perspective, they, you know, they sound pretty engaged about everything. Um, we did 3 sessions, which was a challenge in itself, uh, of applying Sensi and getting it set up and getting the data in and configuring everything, and then detecting, uh, issues, and then figuring out what those, uh, issues with the data correspond to in terms of real failures. That was a real challenge in itself, or even then, um, when I came in and did the final session going through some more complex things like setting up some off-state filters with derived measures, um, there was some really positive feedback in there of, um, good engagement from the students, even if it is quite a tricky concept to get your head around, especially with, uh, say, uh, time windows within the derived measures. And has there been, because obviously we've got gens of AI functionality as well, co-pilot and things like that, has that been introduced yet? Oh yeah, yeah, yeah, yeah, we also covered that in the final session. I believe it was Maisy that took over. Uh, the demonstration of the co-pilot's functionality, but I think that's a, a great functionality within Senssight in terms of being able to provide that context. Um, so it, it's my hope that the students really got some more insight out of that from being able to ask, OK, what does this mean? Can we go into more depth of this and have co-pilot come back with those answers and be able to provide a lot more context surrounding, you know, a data set that is realistically highly synthetic. I mean, that is a really interesting part though, isn't it? When you're having that conversation with a, with a, with an AI about the problem, your specific data, that is just such an eye opener that really isn't it? And I think that definitely that use, the use of natural language in analyzing and understanding technical problems is a is probably part of the roadmap going forward, isn't it? That's the future isn't it really, being able to use natural language to, to analyze and to understand complex problems, it's just a bit incredible, isn't it? Definitely, and the capabilities that that gives us to, you know. Teach people who are less familiar as well is, is vast. Uh, it means that you can bring somebody in that is an expert in these things. They can provide copilot with all the information and context around what's going on. Then you can give it to people who are just learning this from the first time or aren't, don't quite have as much experience, and they can ask co-pilot, which can tell them the context that that expert's put in without always needing that expert on hand to answer all the in-depth technical questions. Cause the whole point of what we do through connected curriculum, as well as giving them exposure to the tool set so they know how to use the tool, but it's not just about. Learning how to push the buttons in a, in a particular in one particular application, it's about understanding the direction of travel that industry is heading in and, and the whole Sensei, and that sort of industrial AI really gives them that view into where it is heading, you know, and that, and that's really important, the sort of more general background of it as well as the specifics of a particular application, so, um, fascinating topic. Yeah, I always say, you know, it's, you know, biased, but yeah, incredibly impressive functionality, and it makes it even accessible to someone like me, you know, who's interpreting graphs and things like that, but then you could just go into a case and see, and it just explains it, even before you get to the co-pilot and can ask questions, interact, you can just see summary straight. Way and linking to other cases. So it's so, so powerful. And I think it links a little bit as well to what you were saying, Harris. It, it, I, what I'd imagine it makes it even more accessible technology like that when, you know, when you're, um, teaching, you know, when it's being taught in the classroom here, you can say, look, look how easy this is to use. You can just talk to it in your natural language and Off you go, kind of thing. It's not as simple as that, but I'm simplifying it, but that I'd imagine it's more accessible. And more importantly, the students can continue learning outside of the classroom because they have access to the platform, they can continue you know, using queries to the system and sort of trying to understand how it works or get, get feedback, so converse with that industrial AI. And again, this is a very current topic in especially in higher education as the whole industry is trying to get its head, get its head around use of AI, ethical use of AI and so on and so forth, but also providing another context, the industrial context into the use of AI and how those technologies that could be used to maybe support students writing essays or presentations and whatnot can also be used to query systems and to converse and understand about the characteristics of, you know, an engineering problem. So again, I think that's sort of showing kind of the extra things that you can do with this technology to students and provide a bit of context. You know how they can can generate value for, for businesses. It also gives those students a, a, a great leg up, doesn't it, in in determine future world of work, cos when they're applying to employers, they can put that on their CV that they've actually used, you know Sensei, and probably many of the companies that they would go to apply to may not be even used it themselves, and it gives them such an advantage, cos not only have they got the skill set, they've got exposure to that, that generic technology, and it, it, it just adds to their employability overall, so it's a, it's a, it's a great, um. A great another little aspect for them, isn't it really in their career. Yeah, and actually that was the question I was going to ask, is it already sort of shifted the mindset of students like, oh wow, my career can actually go off here or here instead, you know, all that inspired, which you'd hope it would have done. Yeah, I would have hoped so. Um, students are currently looking in their careers, but I what I can tell you is off the back of the sessions that we did with Jordan and Maisy and the connect curriculum team, a students decided to shift the project. Towards predictive maintenance using sensi using Industrial-Copilot so the functionality there. So again, the, the, the perspective of the technology they were already doing, um, um, sensing, um, probe probe sensing sort for related project and they've taken that and extended that to into the machine learning and again it was they were inspired by that by an interaction that they have with the next group and Jordan. So that's really great to hear. Yeah, that's good, yeah, exactly what, what we're able to see, that's what, what you guys were here for and those, those kind of results and hopefully you'd see that in the future as well, you know, because, you know, with alumni you'd hear about, oh, I joined this company and then, you know, I told them about my experience with, you know, sensor or other technologies. It doesn't have to be sensor what we're talking about a wide range of technology here. That's absolutely right. And again, we're talking, keep talking about this, digital skills gap. So again, having experience with us, this technology again, just data, but how do you. Understand that data, how do you process that data, how do you get help with accessing that data and understanding it. These are all things that those kind of um emerging technologies can help with. So again, those students hopefully will go into industry and they will become the um. The, the SIF are the people who kind of um make the SEIF changes into sort of digitalization of processes and whatnot and with the kind of the right sort of the kind of technology, so I think setting the mindset in the right sort of um. Way to, to do that. …