BIM IN ASSET MANAGEMENT MONTHLY FORUM 11-02-2026
GEORGE STEVENSON I’m going to give you a quick overview of what we’re all doing. We’ve got a combination of BIM4Housing, which has been going for many years, and BIM In Asset Management which is probably 2 years old. BIM4Housing, we’ve got a number of different working groups which meet every couple of months to decide what information is needed to support the various different activities that people are involved with. And from those the teams determine what information, for example, would construction teams benefit from, if the development team or designers gave them more structured information. And the same with operations from construction.
So we’ve developed a range of different standardised information requirements. These are developed in workstreams which them come 8up with standardised knowledge from the community. And these are all published on the BIM4Housing website. We’ve also led the Golden Thread Taskforce over the last few years, bringing together people who are subject matter experts, manufacturers, designers, asset managers, to really determine what the Golden Thread needs to contain. And that was fed into different people. So we identified that different people need different information for different purposes.
So that’s why classification and taxonomy is so important. And also to be able to interpret the legislation in certain standardised ways. So these are the sorts of resources that we’ve got available that we’ve worked on collaboratively, for example, interpreting the different regulations, we created a Miro board for that. But also looking at this from a point of view of the detailed information about individual asset types. What information do we need to know about fire doors, for example, to ensure that we’re collecting the right information. These have been put together by teams of people who are real experts in fire doors, for example.
And we’ve also produced this publication which is a standardised way of doing fire door inspections. Again this is downloadable from the website and it’s quite a high quality document. It supports Regulation 10, but it goes into a bit more detail. And we’ve just recently done one last year for fire dampers and fire stopping. As a collective we are able to engage with the regulators, so these are some feedback that we’ve had in the past from the government and the Regulator. And what we’re trying to deal with the fact is that landlords are inundated with different things that they’ve got to do.
So it’s a matter of really trying to help them deal with the fact that there’s still a heck of a lot of things that are not being done properly. And that’s where the golden thread work that we’ve been doing fits in. And the implications of this, if it’s not done properly, are starting to come home to roost. Where we’ve got projects that are Gateway 3 have failed. Here we’re seeing the Defective Premises Act coming in. And the most recent one is Mace versus Mitie claim against fire dampers. So what we’ve discovered over the years is individuals within our group are finding that in particular senior directors aren’t aware of their responsibilities under the Building Safety Act.
So that’s one of the things that we’re trying to do with the benefits workshop that I’m going to be going on to shortly. Richard, could you show what we’ve got on the website?
RICHARD FREER The first thing at the top is the monthly forum because it’s such an important event in everything that we do. We’re building a community of asset management experts. How do we do it? As George mentioned our working groups. These working groups identify focused issues which are then passed through to our workstreams. These are task and finish. The idea being that they will have an output that will resolve the issue identified by the working group e.g Awaab’s Law, digital handover, fire strategy. And if you click on one of those you can see the meeting notes, meeting video and a 15 minute podcast that summarises that meeting.
Now if anybody is interested in any of those workstreams, we’re particularly looking at the moment for schools, healthcare, residential and AI and data engineering. So if you’re interested or you're working in the schools area put ‘schools’ into the chat. Likewise for healthcare, residential, AI and data engineering. We have an operations working group which should be relevant to everybody in this meeting. So if you want to join the operations working group put ‘operations’ in the chat. Likewise we have an advisory and consultancy working group. This is for people who work as advisers, consultants.
Moving on, this is the Golden Thread Taskforce. This is a highly focused cross-industry team created to establish the standard and provide the expertise and deliver the solutions to make the Golden Thread a practical, achievable and sustainable reality. If you’re interested in getting involved in that put ‘Taskforce’ in the chat. Now we have the Knowledge Hub. We are the Knowledge Hub, this meeting is part of the Knowledge Hub, this is what we do.
This is Black Box, an interesting initiative. This is case studies on issues. It could be a single asset risk or failure or a fully-fledged incident. If we look here on asset risk we’ve got quite a few in there, when something has failed and how you’ve solved it. So if you’ve got something, whether it’s a big incident that’s occurred where you’re working or that you’re aware of, come on here and report it. Likewise just an asset failure, and that can be a resource for you to see how other people have solved things as well.
Financial impact. This is real world examples of what’s happening if you don’t have the right information. George mentioned Mace and Mitie earlier, we’ve got the London Borough of Barking. These are legal, these are all in the courts right now, recently over the last couple of weeks. This one the East Village isn’t in the courts, but it’s a 432 million fire safety crisis for East Village management. So either add to that or use it as a resource for yourselves to convince potentially your superiors as to why they need better information management. And here we have recommendations for fire critical assets.
GEORGE STEVENSON It’s the examples from the golden thread, I’ll step through that quickly then we can move on to the benefits. Broadly what we’ve got now is quite a wide group of technology companies working together. We’ve got quite a substantial team of people now who are, for example, working on digital handover. And we’ve got the education group and we’ve got residential and we’ve got healthcare. So we’ve got quite a strong community around the golden thread and part of the process that we’re operating.
What we’re trying to do is deal with the fact that most projects are providing at least CAD data and O&M manuals, but the problem is that they’re not properly glued together. So often you’ve got a lot of mismatch which we’re going to be talking about in the automate section later. The principle that we’re operating as a collective group of different software companies and providers is the ability to take the various different information that is being produced by different software and then make it so that it’s glued together in a meaningful way.
And where, for example, information doesn’t get modelled, asset information in particular, that can be picked up through 360 photography.
But this is an example of a really strong solution that’s been put together by JJ Rattigans and the client which is the Education Service and Sodexo the asset managers. So what we’ve got here are finding fire dampers, then being able to tie this back to the product information. So we’ve got the golden thread relationship between the need for a fire damper, what product was selected, where has it been installed. And then we’re also tapping into other information that was collected using OpenSpace by the construction team which are 360 photographs that then the asset teams can use, a fantastic way to tire all of these things together.
And from then being able to connect that back out to any documentation is being held, in this particular case it was gathered in Sharepoint. So the documents are then tied back to the individual assets and also we’ve got then a connection through to the maintenance information. And the maintenance information both in terms of what the documents are in the O&Ms, but also tying it back to SFG20. So the principle that we’re working on is that we’ve got information that is machine readable (because that’s what we insist on). We’ve got collaboration and pulling together, that’s providing a context for where that asset is.
So the fundamental thing is what we’re doing is concentrating on gathering spaces and assets. We can also get activities that have been done for different reasons, like statutory maintenance for example, but also operationally critical maintenance and organisational information. And then again our common role there is spaces and assets. And the same with fire safety, energy consumption and also infrastructure within the building. And again what we need to know is where those things are. So that’s really the principle that we’re trying to operate as far as the golden thread is concerned.
What i’m now going to do is ask, we’ve been doing a lot of work over these last few weeks to deal with the fact that senior managers possibly aren’t understanding the value that all of our information workers are actually doing in this area. Paul Sheedy, could you just explain what we’ve done with your report?
PAUL SHEEDY I looked at all of the documentation etc and the issue that I saw very clearly…when this goes to the C-suite of a corporation we really need the FD and the CEO and the COO etc. to understand this. And they’re not going to sit there looking at BSI standard X, Y, Z. So I put together a way of soft, let’s call it the Ladybird book of understanding the golden thread. What it is for the organisations, repositioning this as why you need to embrace the golden thread, what is the advantage to them and what is the legal implications of not following this.
Just to take it back a little step away from the geekiness of our high level details that we have in the niche areas of interest, and just make this far more consumable for a C-suite. So it is really for everybody more of a marketing solution to bring in to markets exactly why they need to start looking at this, why it’s important, why legally to get onboard. But also the advantage for that building, both in terms of it’s set value by having clearly defined digitalisation of what’s happening in the building and just the risks. And showing that this is the right moral thing to do, but also the right commercial thing to do as well.
So it’s written in a different way to most of these documents, it’s written really as a way to soften. And then the second chapter or piece of the document, it gets a little more detailed. And what we’re going to do then is get right down to the nitty gritty where it then goes out to the building managers and the various engineers inside the building itself. It’s aimed at the very start of the C-suite to just get them onboard and take this as an initiative that they need to jump onboard.
GEORGE STEVENSON So what we’ve been doing over this last month or so is looking at all of the different activities that we carry out as information managers to able to determine what is being done as part of the golden thread, what are the principle activities. And also then look at what the outcomes are. And we’ve tried to put some measurable outcomes against those so that they can be seen as an organisational benefit for all of the work that’s been going on. So what we’d really like people to do is to take a look at what we’ve been doing and then take it on to the next stage.
So to help with that we’ve got Mo from Morta and also Franco Pittoni who is a control expert, to form a little group that we’re going to be working on. Mo, can you show us what you’ve set up for us?
MO SHANA’A I’ve been building off some of the work that Richard, George and Franco have been doing. First there are two key things. One, we have a document effectively in place that summarizes the key benefits that we have. So you can see the key ones are information quality, standardization and auditability cost and reduction across the asset lifecycle, all the way down to automation, integration and portfolio scaleability. If we go down to any of these we then have a range of things. So we have what’s the cost base or the function that’s being affected?
What do they actually want to achieve? How could they get it? What are the financial benefits associated with each of these? What’s the risk exposure reduction? And then how can it actually be delivered? So what’s the sequence of activities that I need in order to achieve that and then a summary of what does this mean on a board level. So we’ve brought this information together but fundamentally we’re starting with this as a bit of a summary for why these things are beneficial.
Over time what we really want to do is to validate all of this information that’s here because that’s in many ways what we’ve put out as a collective group as a first draft. And then we think the way to do that is by drilling a bit deeper. And what I mean by that is twofold. Technically speaking, underlying all of this is a set of activities. So George has had that onscreen, technically have…and one of the things we want feedback on ultimately is the set of activities completely right.
Stakeholder coordination, information capture and classification, digital product data and O&M production, validation and data quality. Information federation integration, digital twin and record of portfolio maintenance and response issues. Work construction compliance and planning and then as-built models surveying and AI findability. Now within ideally what we get to is something that looks in some ways similar to a RIBA plan of work, but for handover and for golden thread related tasks and activities.
So I have the set of activities that I just went through now and ideally what we need to start working towards is first we agree on the set of activities. Then having agreed on the set of activities we can break them down into the set of tasks that belong to every one of these activities. And then for every one of these tasks we can start defining a range of things. And I’ve just put up some examples for what ideally should be defined, but this like what’s my output or a deliverable for a specific task? What’s my responsibility like? So who’s responsible for the task, who is accountable for it, who is a contributor, who is informed?
And then I have some additional columns like what RIBA stages does this specific task, for example, or this activity belong to? Because what we’ll see over time is that there’s a lot of relationships between the tasks and activities of the tasks that are associated with golden thread and RIBA stages. Or the same could be like some people can talk about things like 19659, for example, there’s relationships with that too. Overtime we can also start looking so that we can actually genuinely quantify what we have, we can start going this is the time that something actually takes.
But if we were to use specific technologies, and when I say technologies I’m using the word quite holistically, I don’t just mean software. It could be hardware, it could be anything really. But if we were to adopt these factors, what would that then do to the time if we were to use that? So we’re starting basically to build a framework, in many ways, for how we would capture these tasks. And really the idea becomes having a very granular understanding of the jobs that people have to do in order to deliver on the golden thread. And then to identify inefficiencies that are currently happening and then off the back of tht being able to actually suggest the best ways in which things should be done.
RICHARD FREER Has anybody got any questions on that?
PIERRE VENTER All good, I really like it. One thing I’d like to see in there, we have input documents and production documents. And while they could be branded for particular companies, it’s actually to formalise the data that’s required at each of those tasks. And if we get alignment there that would help considerably.
GEORGE STEVENSON We’ve got a very able volunteer in Franco Pittoni helping with this. The point about this…what I think we want to do is arrange something next week or the week after to actually go down and look at this in some detail. Franco can explain to you what we’re trying to do as far as tasks are concerned.
FRANCO PITTONI It’s a matter of collecting tasks from different people that have different experiences and different needs. And making sure that after the collection exercise has been completed we do a kind of cleaning work. So identifying duplicates and making sure that the definition of these tasks is as clear as it can be. And then we have to find the best method to group and aggregate and summarise these tasks, classifying them in such a way that this exercise is beneficial both from the point of view of a working level person and also for a top manager viewpoint.
ALEX OLDMAN I think it’s really interesting to see how the information is going to be coordinated together. I’ve also just been looking at the morta.io website, that’s a really interesting initiative. So yes, I think it’s a good set of tools, I like what we’re doing here.
JOSEPH KOZELKA Yeah, I’m working on something similar. I’m a BIM manager at MTVH which is a big housing association and I’m tasked with understanding the company and different departments and what information they need and why they need it. For a start for our developments that are being done at the moment, there’s a conflict really because the people in development want to get buildings as cheaply as possible, but the people maintaining he buildings want as much information as they need to maintain the building more efficiently.
And these two things are both about saving money, but they are in conflict with each other. And in understanding other people have tried to gather this information and how they did it is helpful for me to look at. Is there a way of getting a copy of this sheet that I can look at and examine in more detail?
Because in my process I’m putting everything together for a COBie spreadsheet on our EIR, but I’d like to see how other people have tackled the same issues about finding out what information different areas are needed, what they need information for and how to best present that to the people actually running departments and maintaining the buildings or whoever they might be.
MO SHANA’A In large part that’s what we want to discuss. So we actually in many ways want to get to as a group is this is these are the steps that we think someone has to take in order to deliver fundamentally that information.
GEORGE STEVENSON Joseph, if you’ve put ‘benefits’ into the chat we’ll invite you to a meeting. There’s two aspects to this. One is that we’ve got the detailed work that Mo and Franco are going to be working on. But we’ve also got the high level white paper report that Paul Sheedy was just going through. So what we want to do on Paul Sheedy’s work is for people to do peer review on that.
So if you’re interested in having that as a document, which is a really good document to maybe share with your senior people, that would be very helpful if you could say that you’re willing to peer review it. Because it’s our objective to actually get that out within the next couple of weeks.
FRANCO PITTONI I was wondering if it’s possible to get an editable copy of Paul Sheedy’s document because it’s much easier to work on a Word document rather than a PDF.
RICHARD FREER Mo’s Morta, we’ll be looking at getting that up online so that everybody can go online and contribute. George, do you want to go on to automate?
GEORGE STEVENSON What we’ve been doing over this last month or so is looking at the way AI is being used to see if we can have some levels of standardisation. We’ve got a workstream set up, a workstream is a task and finish thing. So we define what we’re trying to get out of it and we have 3 or 4 meetings and then come up with some outputs. So these are the people that currently are part of the AI and data engineering group. If you are doing anything and if you’re interested in this area then please drop ‘AI’ into the chat and we’ll invite you to the next meeting on this on the 19th.
Basically what we’re saying is that there is a risk as far as using AI at the moment in that it’s fantastic but at the moment effectively it’s free. It’s being subsidised by the investors in these big companies. So even if you have got a paid subscription that’s not really covering the cost of the AI, so at some stage it’s going to go up. And in part one of the reasons that we’re struggling with it is that what it’s actually being used to do in most cases is to fix unstructured data.
Now that’s great, but it’s relatively inefficient to do that because people are asking the same questions of the LLM. Whereas what we could do is standardise the way questions are being asked, make them a little bit smarter in terms of how we’re asking them. But the most important thing is to actually record what the outcome is and then for a group of experts to be able to say yeah, that makes sense or no it doesn’t. So rather than working in isolation, what we’re trying to do is to look to bring together a group that we’re calling Automate, Golden Thread Taskforce Automate, to work collaboratively.
So the principle that we’re working on here is to establish a community and a commercial model where we’ve got a small group that are willing to share experiences, test tools out, refine methods and then release them for the wider use for the rest of the Golden Thread Taskforce. If you’re interested in being part of that, we don’t know yet what the commercial model is going to be because nobody really knows how the whole technology platform is going to work.
But I’m going to show you now some of the things that we’ve been doing as Active Plan so that you can then say well, we’d like to buy in to that process and bring to it work that we’re doing. Joseph, for example, is doing stuff, then you might be interested in incorporating what you’re doing in the wider community. Here’s an example that we’ve done with an office project.
JOSEPH KOZELKA I’m working with AI, even though I’m really not meant to be because the company is concerned about security risks. But I found that if I can define a problem I can create software using the assistance of AI to solve the problem.
GEORGE STEVENSON Absolutely. I’ll introduce you to some of the members of the team. We’ve got a whole range of people that are coming at it from different directions. The intention of the Automate group is to bring together people that are competent and experienced and are doing things to actually then come up with something that is properly repeatable and auditable.
One of the challenges that we’ve got in handover, because this is what we’ve been using it for, is that BIM models typically, the data that’s coming BIM models that is put into O&M information often isn’t complete. It’s really the design information rather that the installation information. So therefore often the O&Ms are not complete. So this is one of the tools that we’ve created that will allow us to interrogate a zip file of documents and very quickly identify whether it’s got valid information in it or not.
So this would normally have taken, in this particular case you’ve got 10,000 documents and what we discovered was that there was zip files within zip files within zip files in the O&M handover information. So my colleague Tim, he calls it like a Russian doll, and it’s only when you actually get down and open those up that you find that there’s a problem. What we’re doing here we’ll be showing how we can do this, we’re doing it so that we can take a zip file and it will analyse what’s in the zip file and then very quickly tell you what the format is.
So probably identify where there is incomplete files. That can be done in 10 or 15 minutes rather than spending several days doing it. What we’ve also done is we’ve taken a manual process that we’ve been operating for some time now of doing a snagging audit of O&M information. In this particular case there were 54 work packages, so what we’ve done is gone through and manually….if, for example, there is warranty information missing or test and commissioning information missing, that gets picked up. But this is a manual process and this will take weeks to do for a large project.
So what we’ve then done, and the outcome of that is that we can produce a report which is a dashboarded report against the various different work packages to identify quickly which ones are not complete. But obviously what we’ve then done is taken that and turned the process that we’ve been operating from a manual perspective in to prompts that we can then run using AI to also produce the reports automatically. So we’re taking work that would normally take weeks to actually do and its coming down to certainly days. And by doing that also it means that that information can then get fed back to the supply chain to say this information is not fit for purpose, we need that to be done more effectively.
Our objective with this is also to improve the competency of the supply chain. So to be able to then say if they know that something is going to be checked and people are actually reading this stuff, albeit machine reading, then it means they’re probably going to produce better quality information. And these are some of the practical outcomes of that as to the warranties being voided because the maintenance cycle wasn’t reflecting of what the manufacturer’s information was. And also not having the right spare parts information.
So the process that we’re following is to be able to automate the process of indexing basically a document dump, then extracting each document, the reference tables to run to see if it exists within the file index. And then also checking each of the work package submissions and then also drilling in to the detail of the consistency of naming. Are floors and spaces consistently named, checking the schedules etc. So these are the use cases that we’re following at the moment that we’d like others to join with us on. Because if we do this together we’re going to get something much more interesting.
This is just one example here where we’ve done the analysis, this is the COBie data from the BIM model and you can see here that all of the door numbers are missing. And also the spaces that are in the model don’t reflect the spaces that are in the door schedule. There was also 25 different door types on this project that in fact in the O&Ms there was over 100, so we need to mop those together. So we’ve created the tools to be able to manually extract the information from the PDFs, but really we want this as proper machine-readable information coming through from the supply chain.
And also this is reading the CAD drawings and we’ve identified that some of the doors are on the wrong floor. Here for example, on this single floor we’ve got one that’s on the second floor and one that’s on the first floor, and this is in the same drawing. And this is really because of the way that the CAD data has been created, there’s not adequate care being given to the production of the information because people aren’t expecting people to actually use it. But if you’re using this to do proper relationships with schedules and planning works, it needs to be right.
So what we’ve done is then taken that information and then we’ve made it so that we can then tie that back to the right data, so that we end up with connected ironmongery to that process. So that’s where we’re going with this, also to discourage handover information to be including this type of water marking, because otherwise that’s very difficult to read. Now we understand why they’re doing that because they don’t want door schedules to be used just to go out and procure from a different supplier, so that’s why the ironmongery specialists tend to do this.
But after the information has been ordered and installed the FM team should get machine-readable information, there is no reason not to have that machine-readable information. So that’s probably something that we can do to get the information structured properly from the start. And also being able to read the information from spreadsheets so that we can actually extract the information and then go from there. I’m going to stop there and if anybody is interested in joining that AI group you’d be welcome to join up.
KEVIN PHIPPEN All I was going to say is using the AI tools in that way, I guess what you’re doing is correcting the error, or at least identifying where the errors are and then coming up with a kind of work around in order to make it all useable information. But surely the objective of the AI or the use of the AI ultimately is to make sure you don’t have to do that i.e. to put the process right in the first place, so that the information that’s coming through immediately plugs in to what you’re plugging it into.
GEORGE STEVENSON I’m pleased that you’ve said that Kevin because that’s actually the point that I didn’t emphasis adequately. What we’re trying to do with the Golden Thread Taskforce is get it right first time, that’s what we’re all working towards. But the fact is the industry isn’t doing that at the moment for whatever reason. And I think possibly one of the reasons is that they’re not being held to account when they’re producing incomplete and poor quality information.
KEVIN PHIPPEN I get that, but I guess that my concern would be apart from trying to keep the processes as simple and effective as possible, is that if you base everything around putting lots of people in place in order to fulfil the requirements of the golden thread you create an overhead within the business that probably isn’t sustainable. And at some point it will get chopped back by somebody and then you lose that golden thread.
Over 30 years of a building’s life, that’s an awful big commitment to maintain that data throughout 30 years if you’re having to handball it from one system to another. And you’ve got lots of people involved in that. And I think we’ve seen over the past 20 years that that kind of administrative layer of people that the specialists in middle management have been removed because it was unclear what they were doing or what value they were offering to the business. And I think if you continue to build around people in that sort of golden thread you’ll end up losing the integrity of the data.
I recently did an exercise that i’ll happily share with you that looked at statistical probability. So if you do everything with all of the right numbers of people throughout the whole of the RIBA cycle and assume that a building has the life of about 30 years. And you do everything as immaculately as you can, you’ll end up with a 3 million pound error upon handback in PFI. So that’s doing everything correctly and that bears little comparison to what we see in reality with what we see in reality, with probably a half if not a quarter of the number of people that should be in the loop.
So that’s always my first concern, is making sure that the process is right at the start of the building design because that takes out a lot of the costs during the operational phase later.
JONATHAN HALLAM In my last organisation one of the things which was looking in to AIs was around naming convention, but particularly files in the past which wasn’t named to the standard that was required at that time. So they were looking in to files and picking up the relevant information which were needed for the component of the name. One thing I want to get on to is the fact of the bit which the AI needed was to train to say I want to find the date.
Or I want to find the asset number or the key components which need to go in or, for example, the component which will link with your originator which is in the file name. But the key bit of that is that the forms which are being trained for the AI to ask are all looking the same. And because you’ve got so many different suppliers doing different layouts of forms, even though you may be asking the same question the AI may not pick it up because it’s in a different form template.
Yet actually the thing which we’re all doing, even though they may be from different suppliers, is all the same inspections, the same activities which are being done. To put them all on slightly different formatted forms which you’ve trained the AI to do one thing and then you come across from a different supplier and it’s struggling and you’ve got to train it again.
TIM AIKIN There’s a few things. First of all, if you’re just using the Large Language Model you’re not having to do any training at all. What you’re doing is just altering the guardrails, so that is the instruction sets. And the work that we’ve been doing which started as a research project to do this analysis is we’re checking to see what comments it came up with, what it passed and what it failed, and then looking at the document and going is that valid or whatever. But also part of the thing we’re doing is oh, what would you need to do to make it pass. This is like show me how to cheat, but not quite.
It’s almost like all of the information was there, you just didn’t articulate it. And there are two things, that’s why we frame this as automate, is that some of it is AI interrogation and some of it is actual machine automation, which means that it just does checking. And sometimes you just build a programme. like yesterday I used live coding which was mentioned earlier I think by Joseph which is where he’s using AI to generate the code, help him write the code and then run it and so that you can end up with something.
We’re also doing the analysis on data sets which was to do with help desk. And what I’ve said there is that it’s giving me some information and I’m saying right, can you help me write a Jupiter notebook? Now if you’re a data scientist you’ll know what a Jupiter notebook is, but basically it will create all of the programmable things that will give you the same output. So you’re not using AI there, what you’re doing is you’re using coding to give you those results and export and presentations.
Now when it comes to looking at regulations that’s when you then start to use knowledge bases. And that’s where the AI is simply then acting as the intermediary between the question, understanding it and then picking the right tool to then do the interrogation. It’s not easy to explain in a few minutes, but I do understand that what we used to have to do, for instance in natural language training models we have to create our own dataset and train it on that natural language one.
Now we’re starting to look and say well, I don’t need to do that anymore, I can use the LLM because all it’s doing is interpreting what’s been said, giving it a set of rules to work by and coming back and saying ah, this is what I think you said. And then going off and getting the data. When we’re talking about checking, then we’re talking about two different things. One is looking at black and white so it’s objective and the other is slightly subjective and that’s where we’re looking at content. And then content one is going to be the harder challenge, but just to say did you provide me with any content is the easy one. The hard one is to say well, what you didn’t provide.
KEVIN PHIPPEN Making the point about ontologies and semantics in terms of coding, Tim. To me AI is a bit like a sort of willing puppy, it likes to go out and fetch things as quickly as it can, but you’re kind of chucking a stick for it in woodland at the moment. So it can often come back with a similar looking stick, but it may not come back with exactly the stick that you threw for it. And in order to refine that and make it so that it’s compliance friendly, or at least information you can rely on, you have to put the right ontology in place. You have to put a kind of 3D structure of data in order to make sure that you always hit the right information.
RICHARD FREER So, George is looking for a mentor.
GEORGE STEVENSON What we’re trying to do at the moment…in the past we’ve shown what we’re doing as far as the digital asset capture is concerned. And what we’re now in a situation to do is to bring together people that are willing to get involved with that. So if I just show you a couple of examples of what we’ve got, if you’re interested in either being a mentor for asset capture, or indeed a user for asset capture. And we’re working with some universities as well at the moment to then look at how we can then bring that together.
For example here, this is an office building that we’re working on at the moment where we’ve got the asset list. And everything now has been tied back to a 360 photograph so that when you click on it it will take you to that particular CCTV camera in the building. Now we’ve got a hospital, for example, that’s got 3,500 360 photos, but they’ve not been tagged yet. So we’re working this around, there for example, by clicking on that fire door it takes you to that individual fire door in the building.
So if you’re interested in engaging in what we’re doing here then it would be very interesting to see if you want to participate. We’re planning to have a session possibly every morning at 9 o’clock for half an hour to get people engaged in this to be able to start practicing both as people that are doing tagging, but also as people who can act as a mentor.
So Kevin, you’ve done this type of work before, so it may be that you’d be willing to just oversee 2 or 3 younger people as to whether they’re picking the right things. And we can’t break anything here, it’s a really good method of training. So if you’re interested in the digital asset capture or digital asset verification process then let us know and we can include you.
RICHARD FREER Let me stress to particularly the new people in here, what we like to do is have a brief one-to-one conversation with you. At some point over the next couple of weeks one of the steering committee will be in touch to make an appointment with you for a quick chat, just to see what we can do to help you and what you’re actually looking for from the group.
GEORGE STEVENSON For example, we’re just going to have a chat about retrospective fire strategies which the asset management teams are recognising are particularly important at the moment. The combination of what I’m showing you here with the retrospective fire strategy plan is very, very consistent. Because the 360 photos that have maybe been put in to enable them to monitor this particular access system, that information can also be used to help a fire engineer when they’re looking at a plan they’ll have a better understanding of what that opening is.
So they can use the 360 photos in line with the other information that they have, glued together to enable them to do work without possibly having to make as many visits to site. And that means we end up with a lot of cost thing. Matt, can I call on you to explain that?
MATT WILKIE I sit on an advisory service for an SPV and we manage other accounts for SPVs as well. So our main focus now in the industry is for handback and one of the key issues that we’re having is understanding where most of the estates are because a lot of the obligations sit with the FM companies, they tell you that everything has been fine for 20 years. So as part of the due diligence piece we have to make sure it is actually fine and although the fire strategies in theory don’t actually sit with us, they may sit with the NHS, they may sit with the FM, they may sit with the tenants.
In reality we need them to be correct, so we’re actually doing retrospective fire strategies and that then triggers a load of other events when you start to find variations and stuff that nobody has documented. So you may have a fire strategy that can be 14 years out of date on some estates. And the problem that, for example, the trusts and that have is trying to get all of the tenants to tell them exactly what they’ve been doing, how many people are in the building. Or they don’t update things, so it just gets pushed down the line to next year and next year and next year.
But the implications can be quite significant because when they alter things like door access systems it can impact the site because the people with the access may not be the same people that need to evac in a fire. But now they can’t get out of the building because the doors are blocked off. So you have things like that that do occur on some of the sites and I am actually pressing that quite a lot when variations come in just to say have you checked this? How does that interface with your fire alarm system? How do you update your strategy? A lot of things people may not have asked previously.
But with the retrospective surveys it also then creates some changes that can be beneficial or make things a little bit more difficult. For example, you may find that you can reduce the number of fire dampers, for example, depending on how you rezone the whole areas. You may find things like door access systems have been installed in areas that it shouldn’t be in. You may find that there’s been walls put up which was in an evac corridor but it’s no longer accessible because somebody has decided to make it a help desk or hot desk area and just decided to get rid of the door.
So they do have some significant impacts and the other bit that a lot of people miss is when things move from business as usual to where it can drag in building control. And then it can end up triggering having to bring everything up to the latest standards and not the standards of when the building was built. So there’s a lot of factors that you’ve got to take in when looking at it because there’s a lot of risks that you may have hidden away there.
Assuming that you get the strategy done, everything is online, everything is fine and then you start seeing some trigger events around fire doors that were never certified because of the age of the building, but they altered some of the building, so have they done the fire doors? Is it a 30, is it a 60? Is the wall 30 60? And this starts a ripple effect in opening a can of worms. So they’re important and need to be done, but it’s also understanding that there is a lot of risks of actually doing them correctly because of things that people have done over the years.
It’s not stuff that we’ve done, it’s stuff that we’ve inherited. And that’s the biggest thing is getting people to understand it’s a collaboration, it’s not a finger pointing event. If somebody takes a wall down there is no point us complaining about it 14 years later, we just need to get things sorted now. So it is important that everybody is on the same page and we need to fix it. I'm not interested in who did it, when they did it, why they did it, we just need to get everything where it should be so that we can hand the buildings back fit for purpose.
And that aligns with the Regulation 38 because if a trust, the NHS, a local authority is taking the building back, if these documents aren’t aligned in theory I wouldn’t take the building back because that right liability moves to me. Because you say is everything compliant? Well no, we haven’t updated anything for 20 years. Well no, I’m not signing that off then. So I think that’s where Regulation 38 is an important one because it’s almost like the criteria shifted, the responsibility has shifted for handback.
So it’s not the responsibility of the people that have managed it where before you’d blame the builder and go after the builder. Now once your signature is on there there is a bit of argument space on Regulation 38 which wasn’t there previously. So I think it’s important that a) all of the government entities dealing with handback actually understand the risks and liabilities with this. But b) it’s not one single point of failure because everybody should be doing this and keeping it up to date as part of due diligence regardless, it doesn’t matter which party it is.
KEVIN PHIPPEN Just commenting on the things that you’ve said, I would totally agree with all of those issues in terms of moving walls and being on top of it. But I’d also say that within hospitals especially the fire strategy is something that is very different because it often doesn’t involve evacuating the building, it could mean keeping people in the building and diverting people around them because it might be more risky to move them from intensive care to the outside than it would be to leave them in the building.
So the key thing for me from an asset management perspective are things like how do the air handling systems and vent systems correspond to the fire dampers? Because some of the fire dampers can be left open in the event of a fire so that smoke is routed in a particular way to keep people in the building. And that’s one of the things that I see most frequently in hospitals that have changed where nobody has made a record of that change.
So an air handling unit that used to feed room 1, 2 and 3 suddenly is feeding a different room because it’s been changed as the rooms and the walls have been move around. And I think that can often go on for many years and suddenly you find that the fire strategy is really different to the way it was intended to be.
MATT WILKIE Yeah 100% and I’d also say this is partly why it’s important that people update the RDSs. And when they do the ventilation changes I think it falls under Part F of the British Standards. So you’re supposed to take air changes when you do those changes as well. So you’re 100% right, if a wall is moved, because the ventilation may be in the completely wrong place as well now. I’ve been in schools where all of the supplies and extracts are together.
And I look at it and think wellI know what’s happened there. Whoever was designing this on the CAD has left it all there because he knows he needs to put them somewhere. And hasn’t had the layout done yet because an open atrium, expecting somebody to come back and nobody has ever come back, so somebody has just fitted them all next to each other, so it just doesn’t function.
KEVIN PHIPPEN Well it’s also the fact that all of your maintenance on your fire dampers is done as part of the HU maintenance. So again that needs to be associated with the fire damper that’s looking after particular areas for it to be effective and worthwhile.
MATT WILKIE But also mentioning medical environments. A good one would be physiotherapy or where there is mobility that has changed. So for example if you suddenly put a gym area in where it’s all got hoists and stuff with limited mobility, has the strategy been adjusted for the fact that everybody in that room has less mobility than previously the occupants had.
KEVIN PHIPPEN The other one is numbers. Rooms are often designed to deal with 4 people and when you go in there you find that it’s often dealing with 30 people because the NHS have compressed more and more people in the room and that’s got to have an impact on the fire strategy.
RICHARD FREER I was just going to call Ron about Regulation 38.
RON HEDGER I’ve been putting things in the chat because it seems the easiest way to do it. I don’t have a specific presentation or piece to give today.
RICHARD FREER Let me ask you about what you said about Regulation 38.
RON HEDGER Building Regulation 38 applies to the signing off and completion of works that are controlled under the building regulations. I put that in because Matt, you appeared to be referring to Regulation 38 as an ongoing record of, in this case, the fire precautions within a building over maybe a 20 year period of change. That’s not my appreciation of the case, Regulation 38 is a one shot at the end of the building regulation process to say yes this is complete and the information from this task has been passed over, end of chapter.
And that doesn’t happen through the day to day management and evolution of the use of a building. That’s my appreciation, that’s the whole point of the golden thread is to fill in that space. so the golden thread should record those day by day changes through the fire strategy of a building, whether it’s because they’ve changed the use of a room, added more people to a room, changed the capabilities, decided to put in access control. Anything like that, technically it should go through building control.
Even a change of the type of people that they put in a room is a material change under the building regulations which requires a reconsideration of the fire precautions. But that would mean that building control would be just buried in daily changes if everyone did it according to the letter of the law, so it’s quite reasonable that that’s not the case. But all of those changes must be recorded, otherwise the fire strategy evaporates.
GEORGE STEVENSON Ron, if I could just question that. You are absolutely the expert, that’s why I’ve asked you to be here. But I’ll just question what you’ve just said there because Regulation 38 isn;’t only about receiving information, it’s also the duty of care of the responsible person that’s having work carried out to provide the contractor with the information that they need to ensure that the work is going to be safe. Am I wrong about that?
RON HEDGER No, it is a two way activity, but it only applies to building works that are regulated by the building regulations. So in the absolute black and white legal interpretation then yes, the changing of the number of the people in a room from 4 to 30 is actually building works and therefore should be regulated under the building regulations and should be subject to Regulation 38. But applying the regulations to that level of detail would be untenable. However, the changes have to be recorded.
KEVIN PHIPPEN Ron, can I ask a question? Talking about Matt’s situation where he’s handing a building back at the end of a 3 year contract and that building then becomes the property of someone else. What’s your view on certification of that building prior to its handover?
RON HEDGER As far as I’m aware there is no fire safety certification process. It was actually quite reasonably covered in the previous legislation the Fire Precautions Act, with literally fire certification which was directly enforced by the fire service. That’s no longer the case, there is no building, general purpose work place or other building certification process. it’s my understanding that the Building Safety Act and the Regulator is moving towards building certification for large residential buildings.
But it’s not in place yet, so a higher risk building will eventually have a building safety certificate which will include the recording of the fire strategy, the fire precautions, and associated with that the management and maintenance of those. But that doesn’t exist other than for higher risk buildings and that process is not mature yet.
ANDY COULDRIDGE Ron, can you confirm for me Regulation 38 has to be completed prior to occupation?
RON HEDGER Yes, it should be the process of handing that information over in all necessary directions should occur before the building regulation completion certificate is issued. I have to put a disclaimer in here, I’m not a building control officer, I’m not an expert in the full range of building regulations, only from the point of view of fire safety. But the Regulation 38 is something that should be completed in order for the completion certificate for those regulated building works to be signed off by the relevant building control body.
ANSY COULDRIDGE Then it comes under kind of the Gateway 3 model. Anything that was designed and built under Gateway 2 where Regulation 38 actually isn’t necessarily being handed over, or the information isn’t being handed over or validated. It means that actually occupations are taking place prior to that being completed. But I’m not sure that if at this moment in time the Building Regulator under anything Gateway 2 is required to sign off anything at this moment in time. Gateway 3 is what I’m hoping ad praying is going to make my day job a bit easier, but currently obviously it;’s a bit of a challenge.
RON HEDGER My understanding is that Gateway 3 requires that the design that’s agreed at Gateway 2 and is fixed at Gateway 2 and cannot be varied from, other than through the variation process that they put in place. But that must be demonstrated to have been built exactly to the design that was set at Gateway 2 and Gateway 3 incorporates the handing over of all necessary Regulation 38 information.
ANDY COULDRIDGE Yeah I think one of the things that i’ve raised on previous calls is that dealing with the Regulation 39 handover, the digital handover evidencing the fact that most of what I’m receiving isn’t an as-built document. The other side of that is that when I challenge it, the observation that i’m raising is that it’s not we procured, we didn’t procure an as-built. I think Gateway 3 for me is going to be a bit of a godsend, I’m hoping it’s going to make things easier. But what I haven’t read is that under Gateway 3 the Building Safety Regulator will collect the data, will check that everything is in line with what’s been planned, but not necessarily what’s been built.
So the validation of that data for me still poses a risk, or the risk being that actually what we’ve got is data that is inaccurate because it wasn’t built like that. But what we’ve got is something that represents a drawing. What I want to say is that i’ve recognised that data is great. A BIM model would be hugely helpful because that would then help to validate that, developers currently aren’t going down that route because of the cost associated to it I’m guessing.
But more importantly, when it comes to the onward management of Regulation 38 and the golden thread information, what I’m conscious of is the fact that we’ve got the data that after everything said and done still doesn’t equal what might have actually been built, and that to me is a concern. And i don’t know what the answer is, but I think you can’t beat the good old fashioned actually getting someone to have a look at the cable that runs 27 floors that actually isn’t the right specification cable and it quantifies or equals a mandatory occurrence report because it’s not the right type cable. And that’s a real example.
GEORGE STEVENSON Just to build on that Andy. What I think the practical way forward with that is 360 photographs. Because 360 photographs are very low cost to do and they’re a mechanism to then validate that. So one of the questions that i’ve got for the community, if you know people who are knowledgeable about fire, mechanical services, electrical services, fabric FF&E. If you could make them aware of the fact that we’re looking for people now who have got that expertise, who could act as either being able to identify what a valve looks like and what a distribution board looks like.
And from a competence perspective, but also to be able to train other people up. So these are likely to be fee earning tasks that we want to just connect with people that have got the appropriate knowledge to be able to do it. So if we can reach out to our community of people and say if you’ve got expertise in these areas then please get in touch with us because we’d very much like to talk to you.
RICHARD FREER Just to remind everyone if you’re interested in education put ‘education’ or ‘schools’ in the chat, residential put ‘residential’, healthcare put ‘healthcare’, AI put ‘AI’ in the chat and we’ll be in touch with you.
JOSEPH KOZELKA I was just going to mention what your previous speaker said about wanting an expert to look at what a valve looks like, for example, or what a various fire component might actually look like for their validation under video capture. And again, this is something that comes back to AI. For example, I wanted to bypass the gate where i’m living because we haven’t got fobs for the actual gate system itself. And I took a photograph of the gate and the button that opened and closed the gate.
And then it very quickly identified everything: it told me where it was made, where it was manufactured and who to contact, everything I need to know about that particular thing, it just looked at it and told me what it was. It’s got a control box so I went outside and took a picture of the control box and it came back yeah, that’s such and such a control box manufactured by so and so and somewhere in South London is where they make it, these are the contact details, this is schematic how it works etc. AI just, I was using Grok at the time, I suspect other programmes do the same thing, but that might be the solution if you’re not sure what something is you can just feed that information into AI and it will tell you.
GEORGE STEVENSON Sure, whether that’s coming from a person’s experience or whether it’s actually coming from a technology supported thing, that’s exactly what we want to do. But at the end of the day we’d like to have some people in this process. And if there’s people who are willing and interested in doing asset tagging, for example, we want to be able to train the next generation up so they’ve got the knowledge to be able to do it. We don’t want to be completely dependent on AI, I don’t think.
KEVIN PHIPPEN Just commenting on the point that Joseph made there, if you’re talking about an access gate and you go and take a picture of the access gate and a picture of the control box that’s controlling the magnetic opening of it, it wouldn’t necessarily know as AI that that might be part of a bigger BMS system i.e. controlling all of the access control for the building.
So then if you go and appoint a contractor just to maintain that bit of it or just to investigate a fault with that bit of it and try and override it, you could be creating all sorts of other problems within the overall access control strategy. So I think that’s where AI can be a problem as much as a solution when you’re looking at things in isolation without seeing the bigger picture.
ANDY COULDRIDGE I don’t disagree with you there, but I think in Joseph’s scenario, that type of scenario is when you’re working with suppliers, what you’ve got in the absence of having this kind of digital data, something that tells you what you’ve got (particularly on older buildings perhaps), it kind of gives you something to give to the contractor. And I agree that it might be part of a wider access control system, BMS system or otherwise, but it gives you something to take and give to a supplier in terms of arriving with the right parts.
And I think where there’s value in that is that you kind of arrive with the parts in one visit and not two visits. I need to have a look at it first to see what I need to order and I need to go away and come back. So there are cost implications, or cost savings in that type of scenario which in the resi world, for those of us that are managing residential units when it comes down to cost and service charge and having to explain that away at the end of the year to leaseholders, any saving adds value.
RICHARD FREER Martin, you put in the chat that you can fulfil that role, is that mentor?
MARTIN JORGENSEN What we were just talking about, as far as someone who works in industry to identify things. That’s basically what I do day to day.
RICHARD FREER We’ll be having a meeting again in about 5 weeks, probably around the 18th of March.
CHAT
Steve Jones Support for legacy BAS installations? There may be no data available as a starting point.
Chris Jeffers It would be really useful to have a "deep dive" on use of Morta and what is being developed here
Lawrence Chapman useful if the activity type is classified using the Uniclass Activity table Ac Activities | Uniclass this will allow bechmarking different activity types.
Carl-Magnus von Behr
I think the lack of factual accuracy is also a problem
Raymond McGurk part of our group has a software business happy to get one of my team involved to discuss automate
Jonathan Hallam block zips apart from GIS files where files need to be collated & kept together
Barry Bassnett Presumably, if the information is incomplete, they haven't met the commercial agreement, so you can place rejections or retentions on the payment until the information is complete. Great tool for clients and contractors. Of course, it is key to building safety reports.
Ruben Romero Vivancos but you need computer scientist not BIM or asset managers or document controllers because no capabilities
Lawrence Chapman Critical LLM is trained on fully technically assured information, and AI understands the context of the information container and contents of the information container.
Ashima Sawhney we have just started this process for fire safety items. We have just had one of our high risk buildings undertake a 360 camera
Barry Bassnett Happy to mentor the 360 capture process, how to get the best imagery in tricky lighting, for example, and geolocation in the site plan, where did you actually take that image which can be the bottleneck in the process IT can be straightforward providing you do a couple of things in the field to help s automate the process and how you can document any changes
Ron Hedger I agree, maintenance of the fire precautions is almost impossible without a clear record of what the fire strategy was referring to. The availability of 3D photos, or clear records of what was in place last time the Fire Strategy was updated, allows changes to be identified, and an assessment of the ned for additional measures can follow.
Joseph Kozelka Im not sure I can convince MTVH to survey their 60k buildings
Ron Hedger Regulation 38 only applies to changes which pass through Building Control. general day to day changes have no requirement for Regulation 38. This is what the Golden Thread is supposed to remedy.
Lucy Kean Not forgetting that Reg 38 for HRBs is different as it is Reg 38 of HRB Procedures not building regs
Grant Bezuidenhout George, you hit the nail on the head, we need to ensure we do not become over reliant on any tooling. Tool agnostic is, afterall, the entire focus of "good BIM/IM"
Steve Jones Our products unlock the data from legacy BAS installations and set the stage for transitioning to current generation technologies and complete transition of the building using every concept discussed here.
Steven Anderson I think what we are identifying that this is multifaceted challenge, and the approach needs to from a People, Process, Data and Technology perspective
BIM in Asset Management Monthly Forum – Structured Summary (11 Feb 2026)
Meeting Overview
This forum brought together BIM4Housing, BIM in Asset Management, and Golden Thread Taskforce participants to review progress on standardised information requirements, digital handover, automation, and executive engagement. The session focused on translating technical information‑management activity into measurable organisational value and risk reduction across asset portfolios.
Strategic Context
- Increasing regulatory pressure (Building Safety Act, Gateway regime, Regulation 38 context).
- Ongoing fragmentation of handover information and asset data.
- Need for machine‑readable, auditable asset information.
- Growing use of AI and automation to validate and structure information.
- Need to communicate value to C‑suite stakeholders.
Key Messages
- The Golden Thread must connect assets, spaces, documentation, and maintenance data in machine‑readable form.
- Standardised information requirements and classification are foundational.
- Senior leadership often lacks awareness of legal and commercial implications.
- AI can accelerate validation and analysis but cannot compensate for poor data structure.
- Retrospective fire strategies and handback scenarios reveal significant risk exposure.
- Cross‑industry collaboration is essential to create scalable solutions.
Key Insights
1. Information Quality and Auditability
High‑confidence, structured information reduces disputes, rework, and enforcement risk.
2. Lifecycle Cost and Risk Reduction
Better information reduces re‑surveying, improves maintenance planning, and mitigates handback liabilities.
3. Executive Engagement Gap
C‑suite audiences require simplified, outcome‑focused messaging rather than technical standards.
4. Automation Opportunity
AI‑assisted validation can reduce manual document review from weeks to days while improving supply‑chain accountability.
5. Process Standardisation
A defined taxonomy of Golden Thread activities and tasks is needed to enable benchmarking and automation.
Action Points
- Peer review and finalise C‑suite Golden Thread guidance document.
- Develop and validate a detailed activity and task framework.
- Map tasks to lifecycle stages, responsibilities, and outputs.
- Expand AI and automation workstream participation.
- Pilot digital asset capture and verification using 360 imagery.
- Engage sector‑specific groups (schools, healthcare, residential).
- Quantify financial and risk‑reduction benefits for board‑level messaging.
Implications for Practice
Organisations should prioritise structured information requirements, machine‑readable asset data, and governance frameworks aligned to Golden Thread principles. Early investment in data quality and process standardisation will reduce long‑term operational and compliance risk.
Next Steps
- Convene follow‑up working session on task taxonomy and benefits mapping.
- Circulate draft documents for peer review.
- Identify pilot projects for automation and digital handover validation.
- Continue stakeholder onboarding across sectors.