Assets CollABorative March 2025 - Road Map

IFS Assets CollABorative: Tech Talk - Road Map

Date of Meeting: 11 March 2025 11:00 AM US Eastern Standard Time

Martin Harris Background:

• Director Program Management for Asset Management in R&D
• Working for Martin: 10 Program Managers, 4 UX designers. We are the team that take the inputs that come in from customers through the ideas board usually and develop them into the future product for IFS Cloud.
• We also look after some of the detailed Level 3 level 4 support tasks.
• Been at IFS for three years, but been in around the asset management space for 20 plus years
   
• We are always looking for customers to interact with in a variety of guises initially in terms of requirements and validation of ideas. We typically don't develop things for just one customer. So, if something comes through, we want to talk to as many of you as possible to get your input and validation. Secondly, we want to engage in what we call a cognoscente, which is basically a group of subject matter experts and give them a topic. We may drill down to some questions around how and where we should develop capabilities. Thirdly, usability. We will go through and create mock ups and prototypes, whatever we create and we're very keen to have people engage in that process and then we go through validation and then into early adoption as well. So, at any point if you would like to participate and be involved, just drop me an e-mail martin.harris@ifs.com.

Martin Harris Presentation:

Slide: Top of mind for Asset Management Leaders

• First of all, I want to give you just a quick top of mind and this hasn't really changed massively in the last couple of years since we've been doing this kind of top of mind views of the world as a fairly common theme. Last week I was at an analyst event and re-evaluated these, and they're still relevant and still not surprising. So left to right through these, the application of AI. You'll see if you look at the IFS website or listen to IFS marketing campaigns, we are very heavily into industrial AI and how to apply that in the right context. I'll discuss a bit later. We have a number of use cases we've already released around AI in asset management in 24R2, there's more coming in 25R1. There's some embedded in FMECA. We have a list in asset management of somewhere around 20 use cases that we popped up very quickly out of our heads and there's clearly a lot more out there. So, we are going through the process of evaluating those, validating them, seeing what the usefulness is to you as customers, and seeing how we apply and how we build those out.
• We as the product teams, so asset management projects, manufacturing service, etcetera, we rely on a core technology team. They have a team called AI Services who create the functionality which we then use and put in use cases. But key to that really is, AI is incredibly useful, but we have to understand the right context, the risk profile, the ethics for how we use it as well. So that's where we're going with that. We're using it, for example in real effect, instantly in unlocking information inside the masses of data that we have structured for data, and I'll give you some examples later on.
• Second along APM. Most of our customers are on some asset management maturity journey leading from asset management into asset performance management, and that comes in a few slides time, so I won't go into details on that.
• Every customer is at a different point in that journey, and even inside those customers, they're at different points of the journey, according to the asset and the process class and the maturity and the risk etcetera. We are developing solutions around asset performance management. We already have a pretty good footprint out there, and there's more to come in that domain. I'm going to cover that in more detail in a second.
• Third one, asset investment planning. I guess if you've seen any announcements around IFS recently, you'll be aware we acquired a company called Copperleaf Technologies. They are specialists in the field of asset investment planning. Radiance is used when you're doing major investment planning, and you want to look at strategically where you should spend the money. Some of the feeds and speeds that calculation comes from asset management in terms of how it's performed, how it’s been maintained, how reliable it is, etcetera. We've acquired Copperleaf and created the process currently of embedding their solutions into our portfolio. And from an asset management point of view, building integrations between things like measurement data, we have performance data, the asset hierarchies, etcetera into that.
• On the right hand side, ESG. Environmental, Social and Governance. We are very mindful of the demands of the regulatory communities. We're mindful of the directives coming out of the EU and we are making sure that we have the data sitting inside IFS that we can then report as necessary for those customers who need to report scope 1, 2, 3 compliance or who are just using the data to influence their decisions about energy, times, tariffs etcetera. Inside IFS we have a team purely focused on sustainability. And we in asset management are very tightly integrated that team in terms of the integration, in terms of the data we share between asset management, service management and sustainability.
• Four top of mind topics. We've been discussing with all our customers. We are constantly on the look for the next trend, the next mega trend. For example, we know there are 12 key identified AI disruptors coming up in the next three to six years. We're looking at directions for asset performance management to set the building integrations to AIP with Copperleaf. And we're constantly looking at where all these agreements and guidelines around ESG are going as part of that journey. And then putting that into the context of asset management, making sure we're doing things to meet those needs.

Slide: IFS Cloud Asset Lifecycle Management (ALM)

• This is really around the asset life cycle management. I think probably a lot of you have seen this before. The joy of being inside IFS is we cover the entire thing, so on the left hand side, we have things around the asset design. Some of you I think, have been involved in conversations around 3D integration realisation. How we take data from 3D models, ingest that through design process. We have capabilities inside manufacturing, construction, commissioning, etcetera. And then the seamless handover into the operations maintenance phase where I sit in the right-hand side of that circle.
• Whether you're making strategy, break-fix or reactive or it’s condition based, or predictive, etcetera, that’s where we sit into that picture. But again, we have that core piece in the centre around asset investment planning to help make those senseful strategic decisions moving forward.
• There are things as part of that cycle, the right-hand side underneath is around how we manage shutdowns, recycling, reuse, remanufacturing process, those kinds of things. Inside the right-hand side of the circle we have maintenance planning and scheduling looking at short, medium and long term plans for how you do work, but also looking at how will that work comes back and how it gets reused and tracked, if that's serial assets etcetera. We and IFS have a great story to tell around the whole asset life cycle management which is really key.

Slide: IFS Asset Performance Management

• Asset performance management, as an EAM vendor, you would expect us to have a robust story around asset performance management and we have. This is the journey we have. And you can see this graph in many different ways. You can see it as maturity. We have customers I say on the far left looking at the far right, etcetera. Most customers are clustering around these three, some moving towards 4. There's a few going around predictive etcetera. And this is where Valhalla sits right now.
• We have calendar based, we have everything based on plan maintenance, we've have had maintenance planning and shifting capabilities in place pretty much into day one as you'd expect. But even though that is the least mature of customers, least mature in terms of the maturity for written strategies, we're still investing in there, because we can see areas to apply AI to enhance those processes. So, maintenance planning and scheduling, we're looking at things around simulation. What if scenarios. We're doing work already around how to optimise plans. There's a huge amount of work that still goes into that that space. Moving on, we take data in we have measurement points against the assets and so we want to ingest data. From data Lakes, we want to look at the data coming in, set basic thresholds around mins and maxs, and use that for alarm generation, and then use that alarm generation to create work orders or fault reports.
• On top of that, we have in 23R1 or 23R2, I can't remember exactly, we've released something called Asset Insights. We look at the condition of the assets to give you a health score. Condition is something you can configure. So, for example, if you're looking at concrete beam, you can look at concrete cancer, concrete cracking, spooling, etcetera, and define the conditions. We combine that with the age, give a score which we then visualise through a Power BI chart. So, it's really simple to look at all of the assets by cluster, by site, by location, by whatever slicing and dicing you want to do the data, and very quickly graphically drill down to see where any potential problems could be.
• We are planning enhancements to that right now, the asset health users, two indicators, as I said, condition and age, but we have data inside the system around things like MTBF, MTTR and we will be looking at embedding that as another indicator of support, the asset insights.
• FMECA, I'm going to come to in a second so I won't go into that great detail, but fundamentally how a thing might fail. What happens if it fails? How easy is to see if it fails? What's the probability of failing? And then using that as a strategy to help to find when you do your maintenance tasks. I’ll be covering this later on, so I won't go into that great detail for now.
• We also acquired a company a little while ago called P2. Sitting inside what was called P2, now called ENROI. There's a bit of functionality that allows us to do rules based anomaly detection. So, back here I talked about min and max levels. Under this we have the ability to have a thing called integrity operating window where being set much more granular thresholds and have much different alerting. Take time series data in do stuff with that. And again, we are integrating that into IFS Cloud asset management, such that when you hit a threshold, it'll do triaging for you and decide is it a fault? Report? Is it an NCR? or is it a work order? etcetera.
• And the other nice thing with that tool is you can also run machine learning models on it. So, you can run Python or models. So, for example, there's an open standard around isolation forests. And that isolation forest machine learning model you can use to detect anonymous behaviours. So, if you want to do some Googling later, type into Google, “open standard isolation forest models” and read all about that. And again, using the same data ingestion coming into the ENROI platform, pushing that through the isolation forest model, we can use that for detecting anomaly behaviour. And again, triage that according to NCR for work order.
• The only challenge with that is that is a taught model. Isolation Forest is a taught model. It's something which you have to train. Moving to the right hand side. What we want to do is look at autonomous models, so really allowing the AI to look at the data that's coming in and teaching itself. There's a huge advancement the last couple of years around large language models that support that, and a huge amount of progress made around natural language processing. So, there's an opportunity for us to do more work in that AI driven space, where all you do really is ingest data and then the AI decide what that actually means. There's a whole series of conversations that you'd have around that in terms of, OK, how willing are you to let AI loose? How much do you trust it? How much confidence will you have on it? And then once you've allowed AI to do that kind of anonymous behaviour detection, what were you allowed to do next? Would you allow it simply to create an alert for someone to go and look at? Would you allow it to create work calls automatically? Would you allow reverse IoT so it goes back and change data into the SCADA into the machines? We can go so far with this technology, and it's going to be down to the customers to work out the propensity and the base to deal with the risk associated with learning AI in effect to have a reasonable level control over what happens next.
• For us, our point of direction is the autonomous learning. The autonomous looking at data, understanding what it means and how to apply it. It's around allowing the eye to prescribe what the next actions could be. So, create a work order. It's also then allowing it to prescribe what the work task could be on that work order. It's also allowing the AI to define the plan and the schedule, so that's the first thing that t the engineer knows about the broke wind turbine is that he gets an e-mail to say, pop up the roads to Montreal on Thursday because there's a broken thing, and when you get there, there's going to be a spare part by the side, just go and fit it. And while you're there, let AI describe how to do that. So you can use AI through augmented reality, collaborative reality, etcetera, to drive you through the steps to help with the whole ageing workforce, retired knowledge issues we're seeing across industries.
• In terms of back to reality, we are pretty much here in terms of predictive. We've done everything left to right up to there. We are currently looking at how we deploy those models and how those are delivered. We are in this mode right now. We are investigating art of possible. We do have some tools that are in use already. We're parting some things around natural language processing and LLMs. We are edging our way through how much we let AI loose currently. So again, if that's an interest to you and you want to come with us on that journey, then please let me know. We're very keen to let the customers in that space.
• We have done some initial analysis in terms of the types of customers, and the types of industries, and the region's customers working to look at the ability to deal with the risk and confidence that might come out of AI. So again, reach out to me, I’m very happy to engage in conversations on that.

Slide: Application of Industrial AI to IFS Asset Management

• AI. I thought I'd give you a quick flash of some of the AI use cases and the whiz through this circle fairly quickly.
• We're going to start top right-hand corner 01 Asset Definition. Some of you are aware, we are releasing capability in 2025 which is around the ability to ingest non graphical data from 3D models. I.e. Asset hierarchies and parameters. And then link that to lightweight translator models in IFS Cloud. So, in summary, if you have a 3D model of your HVAC unit for example, you'll be able to ingest the more graphical data, and then the 3D model will get passed up via all desk platform service, then transform into a lightweight model, like you'll then visualise through IFS Cloud. So, embedding what used to be called the forward viewer in IFS Cloud as part of that process.
• The trickiest part of that whole cycle is the data mapping, so we're going to use AI. We have developed a piece of capability using a thing called vectors and embeddings, and that allows us to get a really high confidence scoring on data that comes from the 3D model on graphical data set into IFS Cloud. For example, what your 3D model might call breadth, we might call width. And we'll do that mapping as part of the AI process. Very, very simple.
• The second piece on here around the creating content from images. This is proving more challenging than we expected. We want to be able to take photo captures of metre readings and digital analogue and ingest that and associate those metre readings data into IFS Cloud as measurement points against the assets. And before anyone says, doesn't everyone do that through IoT? The answer to that is no. There are many, many cases where customers don't have IoT in metres and still send people around with phones and clipboards and bits of paper and we're trying to help with that. Speaking frankly, that's proving more challenging than we'd hoped. We've not had a high enough quality resolution language model, so we're still working that one.
• Next one down, creating work orders from scanned documents. This is recognising that a lot of customers employ to third parties to generate reports, go and do inspection work. Those reports are often really, really fantastic sources of great data and often in those reports you'll find things like, we recommend you change this philtre in three months time. Or we recommend you do an inspection in five months time. And quite often that those kind of recommendations get missed. So basically, we've bought a capability where we'll scan those reports in through PDF format. Look those keywords, look for key phrases, and AI will generate a list of recommended work orders to be created. And you can choose, say AI goes ahead and create the work orders for you, or you can just present it as, here's a list of things that we think we found in the report and go validate. Massive time saving, great opportunity to reduce the risk associated with missing those recommendations.
• Work task template creation. We recognise that people use what work task templates quite widely. We also recognise that they are infrequently revalidated. So, we're looking at how we can use AI to help with the revalidation of those work task templates, but also looking at can we use AI to generate new ones? Because if we're seeing repeated work, then we'll come back and look at those things in the future.
• Number 3, maintenance planning and scheduling. We've had MPS for a number of years now. It's really good. The thing it lacks right now is the ability to do what-it-analysis. So, what if I have a shutdown period of six months and I've got 100,000 work orders to do? How many resources do I need? and what type? Or I've got a million work calls to do, 100 resources. How long do I need? Or I've got a million work calls to do. I've got three months, how to optimise my team? So, we don't have that ability to do the automated what-if-scenario type stuff right now, we're looking at building that in the future. As you'd expect, that's really quite complicated. So, we're investigating ways of doing that and there's different tools that we have already internally, and we have options to the other work in that space. So again, that's something of interest, then let me know. We particularly think it is of interest for the shutdown turn around optimization processes we see benefit there, but folks who are doing 30, 60, 90, 120 have made their plans, then maybe there's some stuff in there as well. So again, please reach out if your interested.
• Moving to four, we recognise that customers are using text and voice a lot more. A number of examples where customers just type all the information about the work order and the work done into a single text field and it doesn't really get unlocked. That information's inside there. We're not really using it very well, so we're looking at doing some work to look at those big description fields and again search for key data, and from that see if we can create work orders or work task as needed.
• Same with video. Video's even trickier, and so is voice. So, looking at how to extract information from voice recordings. So again, you can just press the record button on your phone, speak for 10 minutes and there's a whole lot of rich information there that can be unlocked. That's what's Copilot uses on teams. So, if you can take that data, that similar approach and then search for key data and then create work orders, then there's an opportunity for us to do some stuff there.
• Structure data for updates to Copilot stuff, we've done that. That's 24R2. We released capabilities, then FMECA. I’m going to come to that when I finish talking about this.
• And then number five up here, then we're looking at again, FMECA using AI to help populate for FMECA libraries. We're looking at using AI to populate the FMECA, but come to that a bit later. We're looking at recommending strategies as well, so I cover these ones in the FMECA section.
• There's a few of these cases we're looking at. We are always looking for more, so if anything's bring that to mind that, oh, wouldn't it be great if a IFS use AI to do this particular task? Or can we use AI for doing this, then please drop an e-mail, whether it's list, we'll come back to you for validation questions. As I said, we love engagement of customers. That's the best part of our job. So, any thoughts on AI, please just come back to me and let me know.

Slide: 2025 Roadmap

• This is what's in the bag for 2025 in terms of the plan. Again, I won't go through this in massive detail, but I will skip through a few of these things. For those of you who are at the oil and Gas forum, you'll remember I put a chart up I think I got three people put up the same charts around work safety. We recognise there's an opportunity for improving the breadth of our work safety capabilities across IFS.
• So clearly we have instance. We have permits and isolations, but it doesn't necessarily all work at the right level. The right connections. It doesn't all necessarily have the right maintenance relationships, and it doesn't all necessarily link up together, and there are some gaps. So, there's a few things I think we're missing in terms of work safety. We've started working that already. We're doing work currently around the sign off procedures and lockout tag out enhancements. We have a whole bunch of stuff to do around permits and isolations. Work safety I think will be four releases maybe five. We'll see. We want to throw an AI thread through that as well. There's a couple of years, at least, of work for us in there. And it's really interesting to expand that out.
• MPS (Maintenance, Planning & Simulation) I have discussed. There's a couple of things happen there. These are simulation also, the way in which we host that as a service is going to change in coming releases.
• MWO, mobile work order interaction portal. We recognise that MWO, the work order maintenance is a bit lacking behind in terms of capabilities. So, we're doing some work on that in 25. Same with tech portal, there's usability changes. Quite a lot of little ideas come through on the board. There are some inconsistencies with fixing as well to address how that's used.
• It's a very technical thing here around OBD project nations to work tasks. Today, the project nations done for the work order level, going to bring that down for work task level. That's undergoing in 25 as well.
• APM, I've covered some of these things already. We're looking at how we do data collection, data lakes. We're looking at how we do monitoring. We're looking at how visualised data, time series forecasting. So, we'll get above this with our ENROI solution.
• Compatible units and FERC. We're doing some work for North American transmission distribution marketplace where we need to change the way in which we deal with compatible units in IFS Cloud today. So, releasing capabilities for that in 25R1 to help the regulatory compliance reporting. That's the FERC pieces. 25R2 will have enhancements and then 26, we will do some more work on that as well. So, there's an evolution of that capability.
• XDIV I've discussed, so the ability to take and ingest data, 3D data. Again, two screens that go there. The 1st is ingestion of the non-graphical data and the mapping of that to IFS Cloud. And the second stream is around the translation and transformation. There's 3D models into a lightweight representation that we can then visualise to IFS Cloud, and if you've ever used anything similar that 3D model, you can then click on the models. See the hierarchy? Vice versa, you can measure, you can section, you can slice, you can dice etcetera with the 3D pieces.
• Asset Investment planning, number 5, as mentioned, we acquired Copperleaf. So naturally we're going to embed that and there's still some scoping work being done as we speak.
• AI pieces I've discussed already, and I'll show you a couple of those in a second around for the FMECA pieces.
• And on the right hand side we have this thing called Product Excellence. When you submit an idea through the community and please, please, please keep doing those. It's really useful. They don't always get acted on instantly, but we build our backlog based on those things. Those things are like annoyances, so inconsistency, layout, this feature doesn't connect to this feature in quite the right way. If there's small pieces of work that take us a few hours or a few days to do, we put those under what we call the PX bucket for asset management and we'll pick those off when we get a chance. So that PX bucket on the right-hand side is always there every single release. We keep biting off the backlog of things.
• So that's 2025. We have a huge amount of work on in 2025 as always. We always try to do far too much, but we try not to disappoint, and try to move things around and that's why that says 2025 and not specifically 25R1 or 25R2. 25R1 is out in May. We actually had the code freeze for early availability was last week. So we have done most of work for EA. We are actually working on some bits and pieces between EA and GA, so between early availability and general availability and 25R1, and at the same time we're ramping up 25R2.
• We always have more work that we can cope with, which is a good place to be, but sometimes means things don't happen. We do publish the road map and the road map does sometimes change due to unforeseen circumstances we can't get out of.

Slide: FMECA

• On the asset performance management journey, you saw we had FMECA in there. We recognised a need talking to key customers a couple of years ago around how IFS supports the concept of reliability into maintenance or pick your maintenance strategy of choice. We actually already had capabilities in that space. We had a thing called FASN (Failure Analysis Structure Navigator). We had the ability to capture problem causes. We had the ability to do some work in that space, but the feedback from a couple of key customers was it wasn't going detailed enough, and it was still jumping out to 3rd party tools or excel spreadsheets to manage the FMECA process. So, we took a strategic decision to build it ourselves. We worked very closely with a few customers in this domain. A couple of people on John's team are FMECA experts, so we engaged very heavily with them. We've built a cognoscente which included some of the people on this call. They included people from aerospace & defence, included a lot of different insights. We did a huge amount of research into things like SAA SAEJA10, journey 11, which covers of standard displace. We looked at what ISO were doing. ISO14224. And the team, have done a fantastic job in the last couple of releases, producing capability.
• I'm going to give you a quick run through what we have.

Slide: Failure Modes, Effects, and Criticality Analysis (FMECA)

• So, what does FMECA do? As I said at the beginning, it really allows you to identify in the context of the item class and the process class, how something might fail. What happens if it fails, What's the probability fails, how easy can you spot if it's going to fail. What's the criticality. What's the likelihood, etcetera. And it allows you then to make a decision around the maintenance strategy. So, if you find something which is high risk, like to fail, etcetera, you might have a different rate and strategy to something which is lower risk and fails once every 30 years.
• So, what companies typically do is they identify the top 5-10% of the critical assets and they carry out a FMECA on a regular basis to make sure they're adopting the appropriate maintenance strategy. And what we're doing is we're putting all of that ability to do that for FMECA inside IFS Cloud and then drives making strategy straight out of it.
• What that also means when you apply AI is you can take that 5-10% of criticality issues and expand across a much broader portfolio of your assets and do more analysis. And the idea is you in fact you create a library of FMECAs based again on the item process class and you can apply that as you have more and more instances, the library grows and evolves under version control. And again, you can build your own libraries and adjust that as well.

Slide: 24R2 / Intention for 25R1

• I've tried to give you a view of what we have today and what will come tomorrow. So, FMECA, you'll be able to do analysis. We've released in 24R2 and we support that with Copilot. You'll be inside the FMECA and you can create the Copilots and you get information about how this combination item process class has failed in the past, how many times what we did next, what the work order, all the activity on that.
• It focused very much in the first iteration around the planned maintenance in 25R1. We focused on doing things around corrective as well and then working at how to have a library of FMECA templates and then how to present the FMECA information in terms of lobbies. And in fact, what we've done is lobbies, lobby elements and home page widgets. So simple graphical representation where you can see the status of the FMECA through life.
• And then our intention in 25R1, is we're doing this piece around recommendations. So, you can in fact ask AI based on this combination of item process class, based on how it's being maintained in the past, based on the work that's gone through it, the meter readings, etcetera. What recommendation strategy would be the best fit. That always comes with a caveat, as you would expect, it's going to say this is recommended by AI. This isn't necessarily replacing skills or insights of a 30 year season engineer, but it's there as an aid, as a guide. Some people might just accept it and go with it. Others we expect will be a bit more risk averse as we go forward. But we'll see how things pan out.
• That's the current plan.

Slide: Failure Modes, Effects, and Criticality Analysis (FMECA)

• I just wanted to throw in a few screenshots just so you can see what it looks like. The first thing you do is you prepare the FMECA, so you work out again the assets you're going to work on, the combination of the asset, the item process class, you work out who the participants are, because often you might have a team of experts participate in the FMECA and/or the Copilot. You'll look at things like criticality emergencies, you'll build views on the data, you'll build insights into how you’re performing, and then you'll come out at the end of the day with a recommendation that says, based on what we've seen, based on the input from people, from AI, here's who was involved, this is the recommendation for the maintenance strategy. Ideally, as I said, you'd repeat this, so you create versions and revisions of the FMECA based on that combination through time.

Slide: IFS.ai CoPilot for FMECA

• This is how the CoPilot is used. So again, we use that CoPilot integration. The CoPilot is aware of the context in which you're asking the questions. You can look at the structured data you have in terms of the actual information you've already put inside IFS Cloud. You can get unstructured data. So, data from PDFs, maintenance reports etcetera. And you can query that data. We've documented controls that say, please don't load your HR personal files into this because then AI look at that data. We've had to make it obvious there's data formats that you probably shouldn't put in the data lake, but now nothing stops anyone from doing anything particularly daft, unless you start thinking about tagging and sub-tagging. That data is in reference as part of the Copilot to suggest the failing modes and effects, etcetera. And we put limitations in terms of how much data can use, etcetera to support that process moving forward.
• Again, response time very responsive based on our testing, we've loaded quite a lot of documentation. We've generated documentation. We've had test data from the Internet. So down here, you'll see that's where they copied interacts.

Slide: FMECA support for Corrective Maintenance

• In 25R1, we've done the focus on corrective maintenance. We've actually built a thing called a Maintenance Change Requests (MCR). So that sits alongside the FMECA. So as part of the corrective maintenance process, you might have put in place a maintenance change request action to make sure we're tracking those changes that are coming from the corrective work. In fact, MCR we're going to apply to a lot of places across IFS Cloud, but it was developed to really for FMECA, but we've seen a lot of use around it. Part of the output from the FMECA could be correct visions of work on PM actions. It could be to create new work task templates, create new PM programmes. It could be, go and send Martin and John with a hammer and spanner and spare part right away to go and do some work. But the idea is that FMECA now just applies structured directive work.
• We focus on where the impacted groups are, we look at what the actions are. We use a journal tab to record the changes that happen. So, we have complete traceability of those change requests linked to that corrective action, giving us full transparency. We use a non-detection tab inside for FMECA to monitor items with the same combination of item and process class. So, we can use that to identify as part of the FMECA process any early issues. As I said, we have this MCR cycle seen inside there helping us do the evaluation of changes to make sure we're doing it sensibly.

Slide: FMECA Library & Lobbies

• We've built in place there for FMECA tree. We in asset management are the first team inside IFS Cloud to build such a visualisation. If you're used to FMECA before, it's typically based on decision tree, and we've enabled that inside IFS Cloud so you can visually see the logic and structure that goes behind the FMECA. Very excited by that. Very keen to get feedback on how you think that looks and interacts. We've done a lot of usability testing on that, so yeah that's quite important.
• FMECA, otherwise is typically it's quite a text based process. This allows you to graphically represent the hierarchal relationships.
• Also, I mentioned we've done dashboards and widgets, so again we can look at the most used failure modes. We can look at the status of the FMECA because FMECA has a status, in preparation, in execution, in work, expired, etcetera. We can see the status of the maintenance change requests on those dashboards to see how things progressing.

Slide: FMECA Library & Lobbies

• We've built things around homepage widgets so we can see the status of activities. We can see due dates on FMECAs and this is a very visual reputation to say, this is a very critical for me, or this is late and so we can see the status of those things.
• And you can follow up. The same with the maintenance change requests. So again, relate to the FMECA, we can see a centralised view of those MCRs and we can have those stacked in terms of priority and colour code etcetera, according to how customers want to visualise progress.

Slide: IFS.ai FMECA Recommendation from Asset Management Data

• This is the recommendations piece. This is the brand new shiny stuff in 25R1 where we are using AI to analyse the historical data. Look at how many times something's failed, or how many times it's failed in a certain environment. Score those things in terms of the probability based on the critical analysis. Look at how many times you've applied preventive corrective planned condition based strategies and then make recommendations in terms of what to do next. So again, using AI as a guide or an engine to help people through the process of making the right decisions around the strategy.

Questions / Answers / Feedback / Responses:

• Q: I just have a quick question regarding the work management. Will it be for all connection types or certain connection types?
• A: To be confirmed under investigation right now, if you ask me that in about four weeks time, I can give you an answer, but I'll look at the transcript from the call and I'll come back to you with a with a view on that.

• Q: I was just wondering for the AI side of the asset management, there's a lot of data management. A data management scenario again for AI engine on an ERP. What is the performance like?
• A: Great question. We're quite new at AI, but the response time actually if you'd like, I can send you a link to a couple of videos. The response time based on the data that we have available is pretty fantastic. I had a demo of it a couple of weeks ago and I was very happily surprised at how respondent it was. Now, the caveat to that statement is what we don't have is millions of work orders. millions of asset, etcetera. We've got a fairly big data set that'll we will be improving on. We've used AI to generate more data for us to test against. And again, so far, the forms have been very, very good. But bear in mind, we're in early releases. We are planning on doing some benchmarking and giving some indicative suggestions in terms of data volumes and data structures as it progresses. So watch this space. We're on a journey to look at that and look at how we expand the data set to desktop, but so far you can write a question to the Copilot and the response is one to two seconds. It's pretty quick.

Next CollABoratives: 

• TBC 10:00 AM US ET / 15:00 GMT / 16:00 CET
  Combined CollABorative: Tech Talk: An IFS Acquisitions Overview

If you are an IFS Customer and you do not have the next meeting invitation to this CollABorative and would like to join, please click here to fill out the form.