SPEAKER_00 0:00

The engineer only needs to fix one or two components in a system of half a million, but to find which ones takes incredible amount of effort. You know, you can turn all the cars outside into EVs, but it will not have the same impact as controlling buildings intelligently.

SPEAKER_01 0:26

Thanks for coming on. Really looking forward to hearing your views about commercial real estate and in particular BMS and uh smart technology and AI. Um so what I'd like to do is just dive right in, really, and hear about, first of all, who are you, what do you do, and why do you do it?

SPEAKER_00 0:48

Hi James, it's a pleasure to be here. Thanks for having me. I'm Mike Churré, and I'm an innovation engineer, which is sort of a shorthand for please make sense of this and build something that works. So I work for a BMS house called 3MSE, and in that I build analytics software that sits at the field level. And that complicates things a little bit, but I'm trying to make it scalable and deploy it in such a way that we can squeeze performance out of buildings, especially around energy, but also about maintenance and occupant experience as well.

SPEAKER_01 1:38

Just a bit about why do you do it and how you got into this? Because I think you've got a slightly uh interesting route into how you got into it. It's not typical, let's just put it that way.

SPEAKER_00 1:51

So on that topic, do you know every time I go into a plant room, I wonder if I'm the first person with a law degree to go into one. That that I wonder a lot because my journey into uh industrial automation certainly is non-standard. I am, I guess you could say I'm primarily a designer because my pivot away from studying law, which was a result of pretty standard doing well in school, getting into a good university and enjoying that process, but not necessarily seeing it as as fulfilling as something tangible could be. So I always found deep satisfaction in bringing something into the world, manifesting something into reality, tinkering, building. And as such, I thought that I'll risk it, see if I can find a job as a product designer. And that brought me into the medical industry where I worked for a while designing the user experiences and the casings for some sensors used by patients in hospitals and in pharmacies to automatically triage, automatically check in. So I got a lot of exposure very quickly to not just embedded systems, but also embedded systems in a highly regulated, highly complex environment. My first foray into the automation space specifically was through Forestrark, a software company specifically focused on building modules and software for the Tritium Niagara platform, which is quite an important building automation framework, precisely because it's open and it allows an ecosystem of software companies to tweak it and build additional functionality on top of it in any way they see fit. So, through that, through my work as the head of product design for that company, I got a lot of exposure to building modern, feeling modern looking applications specifically in automation frameworks. So at that time it was almost exclusively Niagara. Since my time at 3MSC, that has expanded into a lot of trend, a lot of other building automation frameworks, which all share two very important characteristics. They're as resilient as can be, and because of that, they're incredibly rigid. So since then, my entire job has been trying to find ways of navigating very rigid systems to find strategies and find specific technical solutions to unlocking building performance. I think there's an important paradigm to understand is that so much of our built environment is painfully inefficient. And we chalk it up to necessity, but that is not necessarily the case. And in an energy crisis, in an environment where we're trying to be as sustainable as possible, it really feels like it's a hugely overlooked vertical where a massive proportion of our generated energy is used to leave it on the table.

SPEAKER_01 5:14

And when you're talking, just so everyone's aware, when you're talking about automation systems such as Niagara, Tritium, Trend, these are predominantly industry known as uh building management systems. Typically, when we're talking about building management systems, they are a system which is throughout the building. You've got control panels dotted around the building, and then you've got a PC, typically, if you're lucky, um, hidden away in a very dark room in typically the basement. And then this is where the dark arts of managing the building and managing its energy performance occurs as such, isn't it?

SPEAKER_00 5:55

I think, yeah, you're absolutely right to describe it as the dark arts. It feels very opaque, it feels very inaccessible to not just its operators, but especially occupants in in buildings. But yes, the these and other frameworks like it run the critical systems in a building. So that ranges from the HVAC to security systems to powering to electrical systems to monitoring systems. And these systems have really exploded in complexity over the last 15 years, which is why so many people, kind of like me from tech backgrounds, are finding themselves working in systems that only 20 years ago were basically time switches for boilers and heaters. But with the IoT especially, the Internet of Things absolutely exploding, the amount of devices that these building management systems control have exponentially increased. And that's opened up lots of new demands and an environment with a very clear need for new solutions.

SPEAKER_01 7:06

With that brings challenges as well. Um really big challenges to how you run a building and the level of expertise people now have to have in order to run these systems, understand them, and also then manage them. And I think this is a big part of the work which you're doing at the moment, isn't it, in terms of making that more accessible?

SPEAKER_00 7:33

I think it's almost the opposite of the lack of expertise because the operators of these you know huge gargantuan buildings, you know, our office is based in the city of London. So you know, we're absolutely surrounded, so is yours. So we're absolutely surrounded by buildings that have you know entire villages inside them just dedicated to keep them running, and their maintenance and energy costs are so huge. And the people carrying those tasks out are experts in their field. I think they know building operations tenfold better than I ever will. But there's a gap between where their experience and skill is located and where it's needed, precisely because of this scale of real estate that previously was limited, and we accepted that it was incredibly inefficient, and we kind of can't really do that anymore without well, being fiercely unsustainable and also wasting a lot of money. But the way I see it is operating these buildings takes focused expertise, and people who have spent decades mastering mechanical systems, mastering electrical systems, are too valuable to be sent to trawl through haystacks looking for needle, looking for issues. It's um I see the the analogs between the the P versus NP problem a lot. You're familiar with it? It's a millennium problem where can you it it's it's sort of like traversing a decision tree and from one end of it can you prove the solution from an initial proof? And in practice, what that means in building operations is the engineer only needs to fix one or two components in a system of half a million. But to find which ones they need to fix takes incredible amount of effort. And that's effort that doesn't necessarily need to be spent with the right application of analytics and more importantly, technologies that previously uh in fact, not even previously, for a long time now, have been absolutely omnipresent in IT networks, in asset management, in this is almost rudimentary statistical analysis, but getting to that data within these closed building management environments, within frameworks that don't follow the typical rules that a lot of computer science graduates, for example, leave with, means so many algorithms, tools are not usable where they can make a difference in building management. And I think that's why the BMS companies now have gotten so much more complex. They've gotten a lot better at um IT and specifically crucially at analytics.

SPEAKER_01 11:04

It's a really important point to to get across on that front that these systems typically closed. They offer up also a window into the building, but it's a very limited window. Um and I think there's this historic um viewpoint that what I'm reading on this window, what I'm seeing on this window is truth and is actual. When in in reality, something very different could be happening with the mechanical and electrical plant going on behind the scenes, even though you've got a time schedule which says on your BMS front end that's only supposed to run from X to Y. In actual fact, if something's out of out of wax, something's running outside of the strategy behind the scenes, that might not show up there. And you could be thinking you're doing best practice, but the reality is your building's doing something very different because of something which is happening, which is not visible to the day-to-day user of that technology.

SPEAKER_00 12:16

I think trust in systems is a much broader point than just BMS. But I think in building management or in a lot of industrial automation systems, and in systems that control physical objects, you can see the immediate consequence of a lack of trust in a system. But you're absolutely right, and the gap between design with assumed trust, and then the actual use case scenarios by the time trust has eroded causes huge operational problems. Because, like you say, a lot of the time the operators using the head end, using the building management systems, don't really trust what they're looking at. There's a huge piece of work on validating data, cross-referencing data, and mapping their relationships and using the strength of those relationships as a reference point for identifying anomalies. In fact, anomaly detection is an incredibly broad field within building operations. It's it's something that I worked on at uh at Forest Rock and on a very cool feature there. But it goes far beyond that. And cross-referencing a time schedule with actual energy consumption, with the actual demand, with actual um signals from many different field devices. Are they all actually concurrent? Is a piece of work that is often left unfinished because the system is designed, is designed as specified, it gets a tick. Yes, I press the button, that device switches on and off, but no one really thinks about the fallback mechanisms and the validation mechanisms for what happens when inevitably somewhere in this Swiss watch of you know electromechanical systems, some wires get crossed. So I think when it comes to controls, we have a solution. It just costs time and effort. So finding ways to automatically deploy validation software, to automatically deploy anomaly detections to find these kinds of cases is the only way that I've seen effective in addressing this. Because unlike with what we understand as the software that we use day to day, just you know, apps, web services, it's quite hard to roll out and upgrade. You can't, you know, oh, lovely building. Can we just uh patch 1.1, please, next month? Not really, it's handed off now, it's done. So you almost have to deal with an imperfect system onto which you must continuously apply maintenance and apply, you know, all that tender love and care to keep the systems uh speaking to each other. But more importantly, automatically deploy self-validating algorithms. I think mapping these things inside a building and building the right ontologies for these data structures becomes so key there because it means that um it's even possible to validate and to identify these needles in the haystacks for the operators, for the building runners.

SPEAKER_01 16:06

And you've mentioned a word there, which I know you're pretty passionate about, but I think hopefully you can explain it a bit more what it means, uh, but also why having this as a standard across the industry would be very useful, and that's ontology. Um because let's be honest, that's where some of the mess which we're in is due to. Uh, but for anyone listening and hit watching this, I think it'd be great for you to just explain kind of what what that actually means for them.

SPEAKER_00 16:41

So I think it's useful to think of it in terms of models, right? That when a control system controls a piece of equipment, it it's a it's an opaque box of inputs and outputs, and it's an it's an entity that performs an action. How we understand what that entity is and where it sits within a hierarchy is often enough to describe it for us, where we need to understand it as more than just a contextual system of inputs and outputs that are mapped, but the system doesn't really know what it is. An ontology is a system by which we model the structure of a building in data. It's a data model, fundamentally, and it's a way by which we can standardize how we model equipment and data inside a building. And name it. Well, so the naming becomes even more important because any BMS engineers or anyone who's ever had to work in any um industrial automation system will, I'm sure, relate that a lot of absolute nominal naming conventions kind of depend on what the commissioning engineers had for breakfast a lot of the time. There's there's a lot of inconsistency, and that fundamental inconsistency at the very low primary data layer becomes killer in any kind of automated strategies. If you cannot target entire collections, entire groups of pieces of equipment, you're left with manual work that's more, you know, it resembles the kind of engineering that was happening a hundred years ago. You're going one by one, yes, on a on a laptop, but you're going one by one. You're implementing change equipment by equipment by equipment. When a model is standardized, it can be reusable, it can be scalable, and our industry is kind of only learning that, even though standards have been around for a very long time. Huge shout out to the likes of Project Haystack, which is probably everyone's favorite building ontology. Uh, UDMI is another powerful one. But Project Haystack is a um fascinating way by which a couple of very influential um building analytics and building operations companies came together to try to figure out how we can make systems interoperable, how we can make a system transparent to operators and engineers alike in its function, how we can name things but also tag things in such a way that you can slice the data from your building by meaningful dimensions and automatically build I guess build answers to questions where rarely am I, as a building operator, as someone trying to roll out scalable strategies, am I interested in individual pieces of equipment? I'm interested in collections. I'm interested in my entire primary system. I'm interested in the collective energy consumption of all of my primary systems across a real estate portfolio spanning an entire country. These kinds of unification projects would have taken incredible amounts of effort in the past. Data scientists, people extracting data from individual buildings, this would have been an entire project. But with standardized ontologies, this is something we can automate. This is something where a system can access and more importantly understand the data it is ingesting, both in its relationships in a hierarchy and in a descriptive method by which a system can self-identify the piece of equipment.

SPEAKER_01 21:13

So a good example for anyone listening, if we say a fan core unit, a fan call unit on your first floor in one building could have FCU 101 or something like that. And then it could have a completely it could be a completely different sort of naming convention in another building, and a completely different naming convention in another building. So I think it's it's a really important point to take off this, is um apart from the big, big, big, shiny trophy assets where you might have a dedicated BMS team and engineer on site permanently, generally across the board, that's not the case. You've got a BMS engineer carrying out their PPM inspection how often would you say once a month at at best? If If it's a well managed building, really, yeah. And they've got to go to the BMS, attend site, go to the BMS, clear faults, diagnose problems, clear off their schedule. And that's a lot of work to do. And that they're probably only contracted to be there for that day, for that one particular site, and then they're moving on to the next site the next day.

SPEAKER_00 23:12

I think with the majority of buildings, you're lucky to get a BMS at all, really. In in the grand scheme of things and across all real estate. So I think the big shiny assets are a proving ground for innovative strategies. It's not to glorify our industry in the slightest, like the Formula One buildings. Whereas a lot of those technologies will trickle down into the small and medium-sized buildings. And I think the standardization that underpins everything is the fundamental challenge that we have to overcome in order to allow these, to bring us back to scalable deployments. By standardizing at least the data models by which we describe equipment in a building, what we're able to do is automate deployment of analytics, which means any building, regardless of size, can get those efficiency boosts and those operational boosts that automation systems and having dedicated engineers on site gives you without needing a dedicated engineer on site. A lot of the analytic systems that I build are just ideated by me shadowing and sitting behind some of the most experienced automation engineers in the company as they respond, as they identify issues in buildings, capture that wisdom, turn it into algorithms, turn it into tests, and turn it into scalable analytic software that can tell you not just if this unit is off on its maintenance and optimize those schedules in the big shiny assets, but it can also be deployed at the smaller building and then it goes all the way down into buildings where you get no automation whatsoever. There was a very interesting project which Forrest Rock was involved in a couple of years ago for a very large chain of hospitality venues. Pubs, in short, right? We've heard of a smart pub. I think the logic there was that we spend so much time optimizing those shiny assets. There's actually an awful lot of real estate that goes completely unmanaged, which is experiencing extreme inefficiency. And all that inefficiency is carbon dioxide emitted into the atmosphere. All that inefficiency is energy that is driving our energy prices up. All that inefficiency is costs for operators which are not spent elsewhere. Those are operational costs that should be spent on either improving the facilities, paying the maintenance, making the experience better rather than quite literally setting money on fire. So applying these bits of kit, applying this software, applying these scalable strategies into packaged systems that can be automatically, I take that word with a grain of salt because nothing's really ever automatic with deployments, but something that can be deployed at scale in small to medium-sized buildings is absolutely huge for our national carbon uh targets, really. I think wireless technologies have a huge uh role to play there because I've seen companies that really take um wireless protocols to the next level with their resilience. And what that is enabling is a way to not just deploy strategies to already well-controlled systems, which is a big untapped market as it is, but to buildings that might not have any managed systems at all. The likes of you know LoRaWAN is an absolute absolute game changer in the industry. Just because it allows retrofits, it allows regular buildings to be turned into smart buildings without drilling holes into walls, without rewiring, redesigning the buildings with the mostly frictionless deployments. And this is why everything that I do, I think I try to underpin at the very beginning of a project, you have to think, how am I going to deploy this? And how much of a pain in the rear is it going to be to do a thousand times? And without thinking about the scalability of a solution, of a of a piece of software, of a platform, of a, you know, often it's an on-site um, you know, controller running running specific software, without thinking of how am I going to deploy this a thousand times, the impact of that is always going to be limited.

SPEAKER_01 28:19

I think that brings us quite nicely on to one thing which we've discussed previously, um, which is, and I think I know the how you're going to answer this, but I think if I I still want to ask this question, really, is when we're running buildings, people are running buildings on at the coal face in the buildings. They are there's often a disconnect with the strategy which they're working to, the strategy of the end user or well, the endower of the building. They've got an ESG strategy for that building. There seems to be a big disconnect between that world and the people who are actually on it on that building day to day. The people on that building day to day, a big driver is always going to be tenant comfort because that's going to be where they get their complaints. So how do we close that disconnect? I mean, I I've got my own views on that, but I'd love to hear your views on how you close you close that because it is a real disconnect, which is occur occurring, still occurs, no matter kind of how automated or how brilliant your building is, how shiny it is. If you're going to get a complaint from a tenant, the likelihood is you're going to get a knee-jerk reaction to solve that problem, which might have a big effect, a big impact on what the end strategy is for that building.

SPEAKER_00 29:40

I think you'll see this in almost every managed real estate operation because these teams have different incentive strategies. They have almost competing objectives. And I think the solution, and um I've I've actually last week I was I was at a meeting with representatives from three separate teams, exactly um analogous to what you're describing: the ESG team, the energy team, and the tenant team. And all three had an interest in the system being highly effective for their specific end goal. Now, how can we tell a story of what the system will do in such a way where every single member of that consortium feels like this is the right tool for me? And I think the way that we design the systems are often very open-ended. They're often, you can control it anywhere you like. It's it's open. Go wild, go nuts. But when instead of that you design specific features and specific use cases for those teams, for those decision makers and those stakeholders, the same system wears three different hats and works three different jobs for three different user types. And I think although those things are conflicting as knee-jerk reactions, there is the optimum point at which both sides are absolutely satisfied. The tenants need to be comfortable, the building needs to be run effectively, but also efficiently. Neither side must mandate extremes in their direction. And one side wants reliability, the other side wants efficiency. Neither should come at the expense of the other. And by showcasing the inner workings of a system, I often find the teams and the stakeholders with specific angsts, with specific anxieties of what might this do to this sort of to my end goal. Sometimes those can be alleviated, or at least the specific factors that need to be tuned are identified. When you explain exactly what the system will do, suddenly that understanding gives the ESG team the confidence to see, okay, you know, my remit cannot conflict with the occupant experience team. We can precisely pinpoint the control strategy for that building that maximizes one without sacrificing the other, maximizing the uh drive towards those ESG goals while still giving all the occupants a healthy and productive environment. I think by showcasing the stages at which I'll I'll give you an example. Yeah. A uh demand-driven um HVAC system that's uh 3MC actually won an energy award for at the BCIA this year. So I'm quite pleased with that. Thank you very much. But this system uses air quality sensors and occupancy sensors to modulate the um systems running on a floor. But the real trick and the reason why it's been as effective as it is, is that there's actually quite a lot of very specific parameters that are tuned so that the occupants are as happy and as healthy as they possibly can be, while the systems are running at the most efficient. So the magic really isn't in the idea of um demand-driven control. The magic is in identifying the critical tuning factors that makes the system desirable by tenants and loved by operators because it slashes energy use by you know 40-50%. But by putting air quality dashboards on the floors, by placing more air quality sensors than a value engineer consultant would have possibly liked, we were able to really bring the confidence in the system, bring that transparency that there is a system that is doing clever things here. Here is what it's doing. That builds trust. That builds trust in the system, not just because I can go and validate that the system is doing the right thing, but also the act of transparency alone builds a level of trust with different user groups, the occupants, the ESG, and the operations teams that allows them to themselves feel and the occupiers. Absolutely, the the tenants are always at the um you know front and center of this.

SPEAKER_01 35:05

I think I think where you're coming from is essentially traditionally, these were very and I I I can't stand this terminology, but siloed operations and siloed approaches, and people have their own agendas. Now everyone's starting to come together a bit more, and data sharing is becoming more, and you've got to, if you want to take your building on the journey to net zero, you can't just do that in in isolation as a landlord and the uh operations team. You've got to bring the tenant on that journey with you. There's 70% of the energy in that building, right? So if you can bring them on that journey with you and get there by and show them the data, show them what's going on, show them how you're improving it, but show them the opportunities which are there to go further with them. I think that's where the power comes, is bringing all these different stakeholders in these buildings together and having that transparency which you're talking about. That's a change and an innovation which you're talking about. And I think big picture-wise, why is change required and so important in the industry from your your viewpoint?

SPEAKER_00 36:17

Why is change required? Yeah. I think the starting point has got to be that London's commercial real estate used, according to that famous report that everyone cites, me included, you know, north of forty terawatt hours of energy. That amounts to such a huge sustainability impact, such a huge environmental impact, and such a huge cost to us as a society. And not to get too self-important about this, but we do kind of hold the key to that. You know, it's our industry that is controlling the systems using all that energy. The requirement for energy management systems has become so important precisely because so much of this does lead us back to energy use. So much of this leads us to that fundamental obligation to not run these systems at 100% all of the time, if nothing else, just on account of the environmental impact that that has. Because the scale, the absolute numbers that commercial real estate consumes compared to residential, compared to the systems that you know we're used to, we're used to turning things down in our houses, we're used to things using only as much energy as they're needed. That is not the reality with commercial real estate. And if we don't address it, either the entire sector will suffer or the environment will, and neither should and neither can. So that's where I see that need for change to be most acute. Because it's the most unlikely hands that ended up with the power to really drive our country's net zero goals and ambitions.

SPEAKER_01 38:28

It's so much focused on EVs and vehicles, etc. But actually the big how much impact and how much impact we as an industry can make in terms of the country's approach to getting to net zero is is phenomenal.

SPEAKER_00 38:44

Absolutely. Uh you can fact check this for me, but 12% emissions by automotive, by transport. You know, you can turn all the cars outside into EVs. It will not have the same impact as controlling buildings intelligently. It is such an overlooked energy consumer. And it's also an industry where the change really isn't painful. We can make this sustainability change, we can impact this, you know, terror what our energy consumer in quite a significant way with very little societal cost. You know, all of us office workers will still show up to lovely officers, but those officers will not be melting our ice gaps. There's a couple of stages in between, but the point still stands that as a short-term immediate change that is required and necessary to drive the sustainability objectives that have been set, it needs to be buildings. It needs to be real estate. We're not even talking about residential here. You know, 50% of the energy used in buildings isn't commercial. It's all the shops, it's all of the small to medium-sized buildings, like you're saying. If we can design automation solutions that legitimately are resilient, reliable, easy to deploy, scale them throughout the buildings and the country, we will have been on such a gigantic leap forward in terms of achieving those net zero objectives that we're trying so hard to do by buying Teslas?

SPEAKER_01 40:39

Yep. So optimize, sort it out, reduce the load, and then we look at the the renewables. Um so that brings me on to the closing question, um, which I like to ask my guests, which um does commercial real estate need to see evolution or revolution?

SPEAKER_00 41:00

Ooh, I think revolutions are painful. I think you'll have a much harder time finding buy-in for a revolution than for evolution. And I was actually reading an interesting uh article on this: that as a change-making tool, revolutions, historically speaking, aren't the most effective one. They are the most headline-grabbing one, and and everyone tunes in to a coup d'etat somewhere. But incremental change has achieved more, according to this article at least, but incremental change has achieved more to change political systems, or in in this case it'll be, you know, an industry system, but has it achieved more to change the status quo than drastic rapid revolution? I know that's a bit of a centrist sitting-on the fence thing to say, but I think guiding change to where we want it to be will be the safest way to not break eggs along the way. I think that's that's the only way that we can really bring people along with us rather than just mandate and and force change. Because instinctively we will always reject that, and we will always reject mandates, and we will always instinctively be skeptical of you know the revolutionaries can achieve their uh goals, but the rest of us have to play along too. So that's why by making this journey as inviting and as open for everyone, as insightful for the person for whom this is invisible, the the occupant, the the people for who don't really think about our industry much, really. Yeah, but I think by making it something that everyone can see, observe, and you know, play along with, we can make much bigger change than through a revolution.

SPEAKER_01 43:17

Yeah. Do you think, going back to a point which you made and links in with this, do you think that outside of the industry, uh within the industry, everybody knows this 40% um 40% energy um stat, right? I think everyone's very clear on that. Do you think outside of the commercial real estate industry, is that known? And would revolution be demanded by people outside the industry if they knew the impact that this part of the this this industry was having?

SPEAKER_00 43:54

I think that brings us back to our point about transparency and visibility throughout systems and throughout these buildings, because I don't think many people are aware because it's not visible.

SPEAKER_01 44:06

Sammy, you're in a room, you're not in industry. Were you were you aware that 40%?

unknown 44:14

I didn't deny that.

SPEAKER_00 44:15

I think a lot of the buildings that we use buildings have such a long life cycle that any changes that we do to how we design them, they'll have a great impact on our grandkids. Yeah, but we have to squeeze performance out of our existing building stock, which was designed in a time when we weren't really so concerned about it. And I think that's why these scalable control and automation solutions have to be the key to actually getting that number significantly down from 40%.

SPEAKER_01 44:48

Brilliant. Well, thank you so much for joining me today on um Build to Perform. It's been incredibly insightful, and I just I love hearing your passion and your your intrigue into the industry as a whole, but in in particular um automation and the dark arts of the BMS world. And I'm I know we didn't really get into the AI, but I I know your AI overlay passion, which sits there, and how useful that's going to be in the now and into the near future as well. So I think you're doing some really exciting innovative stuff um currently, and uh we can't wait to see where you take that. Um, but yeah, thank you so much for joining us.

SPEAKER_00 45:28

Thanks, James. It was a pleasure to be on Build to Performing.