[00:00:04]
>> But I appreciate you being here I'm really humbled with the opportunity and I'm on earth through to be part of the team you know there are some of the young faces we have in the college that I consider myself now all being here too old 12 years but getting older so that's I guess that's Ok but I couldn't wish for a better team to help me on that and everything good is there if you have any comments you know just address it to me but to be.

[00:00:34]
Let me do a brief introduction. I'll start with. Floyd to me she's a faculty in industrial design with expertise around human centered human centric design and. Chris Bartlett from g.d.r. I would help her explain one of the really interesting applications of data analytics and especially related to the operation and maintenance side of the house and how that information can be used for more informed decision making on the design side and they have a very good project with children Healthcare of Atlanta that you would I'm sure you would find it.

[00:01:16]
Interesting and. Dr Vereker is not a new hire relatively new hire in the School of Architecture expert in building science especially energy efficient building sustainable buildings and so stain able to see through so he will be talking about a scale of data analytics so with the school of building construction here so I'm going to talk about anality accent some of the application of that in construction again thanks for being here this morning I have started our talk with a question that has started my quantitative.

[00:01:51]
Methods class so I see a couple of students from that class here too I always ask What is the decision and we had the decision an allocation the title What is the decision on what you think. What would be a decision. You know there is no longer answer in the morning.

[00:02:15]
You know police. Yeah it's basically choosing but at the essence of that. Are when we are committing some resources we going from one estate to another. I always use Ron Hubbard from Stanford University's the decision scientists decision analysts electrical engineers but it turns to be a decision scientists So he's saying you know I mean the fact that you are here you decided to be here so hopefully this one hour would be good you know this is our job to to make it the best things that could be why do you want to be allocated that and you never got that back at least that's $1.00 that's 3 sources which is top of that this time but this typically involve cost and other issues around that so we all have heard and Alex this you know it's been rising in the cab eluding in the literature from probably around 2000 from May 2000 and now pretty much everybody talking about that in different fields so that's really going to give you some examples of that what what you think about and Alex what it means to you what.

[00:03:26]
Was your understanding of that. Again knowing the answers but still me. You know. Now you want to understand divided on concrete so you want to understand the sense of that so I have a masters in operations research in early 2000 and here so anality x. was used as this sort of.

[00:03:56]
A synonym for operations research but today we have sort of this analogy it's evolved to a field that waits self is called a domain but you know it used to be if you go to the we keep. Web server tabulated they were using the interchangeably so it's a punch of tools and met towards that how we can use data how we can use that every dances and patterns that we can find to bed they're making decisions and typically one of these for for things we either want to describe this situation so we have this script of anality So we want to say you know these are the patterns we are observing So let's identify those.

[00:04:38]
And then we want to analyze that the reasons behind the if the diagnostic part of that I mean sort of I'm getting to the maybe called Fields going from the symptoms to the sources of the problem what is the issue and then using that knowledge because we'd like to predict the future we'd like to have some ideas of what would happen under certain circumstances so that gets us to the next layer which is predictive and.

[00:05:04]
Finally we are all interested in command being the best course of action optimizing a good solution so that's getting you through the prescriptive side of that so the team for sort of I like this kind of life from the Gartner So you know you understand what's happening and why is it happening and how likely that is happening again under certain circumstances and what we should do what is the best course of action so for this thing field this thing type of application we re cannot really do one we felt there are there so information we get we gather here use for better production and better optimization of resources and it was kind of thing so in other terms that we're going to use that a little bit to the big data what it means to you what it means to the practice of architecture or engineering design or city planning industrial design what it means to you Big Data.

[00:06:07]
You elaborate a little bit more. Exactly. We don't talk too much today about cyber security but and also the privacy and other things but we had the back of our mind actually this morning we had that conversation because the project they had with children Healthcare of Atlanta a lot of videos a lot of images so who would have access to that.

[00:06:43]
I would say one thing is is context specific so I see the last building College of computing they have the I mean the Center for big data they have a center for the Southern states you know and now analyzing the big data but this really is specific to the field we are talking about.

[00:07:04]
Volume of course big means that the volume that will last the data are we talking about these 3 features but it's really context specific and that's why you know we need somebody we need then architects to make that decision what is Big Data means for them so it's really context specific dust like about this the Phoenician each sector whenever they get to the point that the existing tools existing methods do not worked out to March and they need something girls.

[00:07:36]
Because of one of the reasons you know either volume velocity or variety of the data then we are talking about analytics and big data analytics so whenever you get to that point that. Your spreadsheet has started through you know crash orders you really cannot get to the point that manage all sorts of things and I'm I get through.

[00:07:59]
2 main things that are interested in this stuff and what if you're passionate about this stuff in the 1st thing you know and these are sort of fundamental and maybe a little bit you know philosophy called theoretical interpretation of the big data analytics one thing is really want to know we have tools that we didn't have 10 years ago 20 years ago and that would continue.

[00:08:23]
Into the future so we can solve certain problems and I give you one example the the whole beige in the statistics it was by a paper around 763 or something like that by the whole thing the whole studio may know the conditional probability that was that although you know almost 200 something years old.

[00:08:46]
But it never got that much attention until recently that we can do a conditional probability assessment. Borking with large matrices of probabilities and conditional assessment that's the whole thing your brain pretty much a lot of machine learning boils down to death in which is about 25300 years old so a new lens to a study of problems and the other thing is new insight I don't want to get too much myself into trouble today but I'm already getting there are some dimensions as humans we can not capture.

[00:09:23]
So we'll be talking about some of these heathen patterns mean we might be some of us might be good into the I mean the architects in the room certainly you are good to have vision for 3 d. some of you might also consider the dimension of the time but we're talking about multi-dimensional and also monetary value so a lot of things evolving and time passes through so how to capture that time gap how to capture multiple valuables and that there might be some patterns that we don't notice and thus the job of data and data analysis to to help so it's a combination of the 2 of them you know we're going to talk about that in our research today so back to our industrial as far as the doctor and you see different application of that today I'm mainly talking about.

[00:10:15]
Numbers which is highly popular that close all industrial But what I mean does the especially we get a lot of images you know we would have the the following presentation after me they're talking about a lot of images I mean we do a lot of. His entire ph d. is looking at how we can make sense of.

[00:10:37]
The financial reports of this big infrastructure of companies can we learn anything can be quantified of risk can be. Use that as a leading indicator of future market so entirely text and numbers but we have a. Deal which becomes more and more popular and the challenge is how we can fuse these data sources and make some sense out of that so.

[00:11:05]
We are going to talk about 3 applications I'm going to stay with the construction side of that and then I'll pass it to talk about energy building science part that we've done how analogous improve building energy stimulation and building energy assessment in general and then we get back to the design human center if there's on with a good application from health care.

[00:11:28]
Chris going to talk to us about so I've been working on a now at 6 for the past 1012 years but primarily I decided to a stick with the cost you know some application we have done on the productivity economy can Alice's but I'm going to stick with the cost because it shows sort of you know how I progress over time and of course in a lot of.

[00:11:53]
Students who are working my group helped me to achieve that so there. I guess you have the all their version of. But anyhow. This was the whole big. Big out the technology is enablers of you know before I get back to my I think at the lowest level we have the digital technologies to capture this data so there are drones or sensors or any of scanning technologies so you get the digital infrastructure in place and then they have to be integrated into some software platform you'll be flat form for instance and then you're going to use that for some sort of analysis and big data analytics comes out the you know we have a lot of your technologies that we can utilize in addition to that so there is a challenge in industrious for us to acknowledge.

[00:12:46]
Integration the vertical side of that going from function to application and of course you know the point you know the cybersecurity you know that scene or what level of access we want to have when we're dealing with deathly So the question from the. Getting back to the construction analysis explore.

[00:13:07]
The challenge that the contractor and owners of major capital have you know whether I would receive enough be there on this job what level of costs they would propose what level of builds that were proposed and if you are beating you uncertain environment how much you have to put down.

[00:13:25]
Definitely you don't want to go to law you get the job but you might leave a lot of money on the table and for the owner to use usually not good because if you have all of these based on that probably that contract allows incentives to compromise quality on your job because they were very thin on their margins and be the analysis you know whether they're making any sense or not and these are really challenges because you know just a lot of contractors play a good rule and kind of try to come up with something that helps them in terms of cash flow but it may not be the best for the project so you have to take that through some the Smart be the analysis techniques and this is something we like you know a lot of projects you you don't have to wait until the project and you know that projects run into trouble.

[00:14:20]
So how we can quantify those leading indicators how we can identify those and make corrective actions so. I talked to many attorneys who work in this field and they said they knew this these things would happen they should have made some corrective actions changing some players in the team but they did not and now we're here in arbitration or other form of legal matters so knowing that you know there.

[00:14:46]
Is there variability underestimation about overestimation is about I mean if you're going to stimulate it's not accurate this to me if you're just locking your financial resources so rather than doing 2 projects you only do one which is not good you want to utilize your resources as effectively as possible examples from a state of Georgia if you have changes over time so this is the volatility of the big prices you need proof price for asphalt line items in all of us that over time and then across the states you know this is the hottest spot analysis for a state so there's a specialty No the down in Savannah the prices tends to be higher than metro Atlanta.

[00:15:35]
So this is a descriptive finality ex-parte of the reasons rerun some issues and you know that the access to supply chain access to the labor access to good construction managers in these parts are different there we get a lot of granite singing from North Georgia from blue to North Carolina and those areas so the transportation costs contribute to that and again it's not just one factor but things like that shows you Ok what what is the problem what are the likely hypothesis to explain this problem so this is again another example of you know from the state of Texas so it's not it's not unique to Georgia so we have the volatility variability all the time so if you just want to use the inflation guess what you know.

[00:16:26]
If you want to add that you know 2 or 3 percent which is most of the estimates get updated No it doesn't for human you are losing some points so we came up with a. Time series analysis that beats the prediction of and forecast of those who consider themselves subject matter experts so this is a genuine poverty of anality so you can come up with the Met towards death or better than what the subject matter experts tells you intuition and their judgment there is a place for doubt but you can't beat that you can take that information and go beyond.

[00:17:06]
But the other question is am I missing the other valuables am I missing other sources of information. So the wide range of data there is throughout industrial from dogs to some 3 sources from census labor. Statistics an economy can now assist so they're giving us good things so things that happens in the market Holly can help me to prepare to bear their estimates their costs for my own projects so there are a lot of things happening I want to know what part of that is a political will to me so again the challenge is these things evolve over time and there are multiple factors and there's a time gap so you have to concede that it does so you need to rigorous statistical framework around that because the sure you have different trends in the housing market shows this trend the cost general costs a lot of lead going up and this is sort of a stumble around around the architectural dealing to index which has been can see that a leading indicator of construction activity so how you can make sense how you put all of them together to make some sense so that was you know the application of multivalued time series analysis which is basically economy tricks so economists extend help us to get better things so we identify leaving.

[00:18:26]
So you if you want to understand the construction costs of course we are still for the most part an oil based economy you have to consider the trend of oil price you have to can see the what's happening in your house in the market. What's happened as far as you know the building permits so 2 days there are a lot of you know there's 3 majors out there don't provide that information they don't use that they they take the not rigorously analyze that we've researched to see how you can take advantage of that and that's a problem of anality.

[00:19:03]
And longer range forecasting so now we get to the artificial intelligence a lot of a structured model is working world when you have short period of forecasting but if you want in the long term so the ensemble machine learning algorithms or some others you know Woodward can really build for the long term forecasting 527210 years as you update your capital program.

[00:19:28]
And I we have another application of diagnostic here which is you know you can detect when the prices changes like this case things that happens after hurricane escapes Hurricane Katrina really find out what happens in the Gold Zone region in terms of. The price. And finally you can do policy implications you know if we find out you know the materials price rescale location would not be that effective you know we analyzed $2.00 sets of data one with that policy one we felt that policy and can see that all controlled factors and that was a notable results and counter-intuitive you know again and what I'm trying to say you know we have to find a good common ground and that's what I'm a striving to on a daily basis I've got leasing the cost how we can work with the capital program delivery to understand Ok there are other sources of information how we can use that thing without them a little bit over my times.

[00:20:26]
I'm going to pass it through a good application into building science energy. Thank you so much for the introduction and thank you so much for Christen later as well we've been working on this for the past couple of weeks it's been a very interesting discussion that I'm very glad that we're bringing to you my goal is to actually talk to you in about 5 between 5 to 7 minutes about one project that is exciting from a building science perspective it's really a pressure to be representing the school of architecture I just joined this past January and part of the high performance building lab we're working on things such as aerial analytics which is the use of drones and building performance inspection but I'll be talking to you today about urban building energy modeling and specifically how can we use data analytics as a process to calibrate such models but I wanted to give you a little story in the beginning this is a special issue of journal called Tad technology architecture and design it comes from the a.c.s. is the architecture collegiate association basically and they had a special issue called measured and it was focusing on data specifically and I applied to it for a peer reviewed journal publication and to my surprise when it was published I was at Syracuse University at the time the 1st article is actually called Everything that can be measured will be measured and it was by none other than my school chair Scott marble who has a really beautiful introductory sentence here called data can be a source of creativity a way to enhance the understanding of the complexity of how architecture operates the world and I thought this was a beautiful way to start how we can relate data to architecture but also to my surprise a 2nd article was also invited by Dennis Sheldon It was called cyber physical systems and the built environment and then is the director of the building lab and it was really great to read what people are drawn to take in the school of architecture are thinking about data so the potential of cyber physical environment is continuous systems fluidity connected across material to.

[00:22:26]
Important conceptual boundaries and I thought this was great because when the article was published. My article was published I knew immediately I'm in the right place because 2 of the leadership is basically publishing in that journal invited to talk about measured data I'm taking more of a and applied perspective so I'm going to be talking to you about how we use these formats of data to understand how people behave in buildings and how we can change their behavior and design through informing their decisions so the context of the study here is that it was a collaboration with a community called Muller and Austin Texas modder is an affluent community that was very interested in energy efficiency for the sake of energy efficiency not really cost savings they have some money so it's Ok they're plastering their roofs with p.b.s. so they're interested in understanding how they can be much more energy efficient it's a New Urbanist community about $300.00 buildings 50 of them were really really rigorously measured so they have these devices gauges and smart meters that are basically allowing them to have access to minute scale data for their appliances This is great it looks wonderful right but no one understands I don't think we would understand what's going on here it's data that is flowing around but we don't really it doesn't really inform our decision so the ideas of informing your decision have become popular through the Internet publicized lists that are telling you what to do to be more energy efficient but it's not really tailored to that community so these lists are often not accurate and these conventional gadgets did not demonstrate a significant impact on the users so we formed a team that was basically focusing on how can we understand the data analyze it visualize it and simulate and forecast future perspectives to basically inform both design decision and behavioral changes.

[00:24:15]
Through the data analytics that we would be pursuing and my focus was on urban building energy modeling specifically the idea is we have some measured data from that community. We want to use some machine learning techniques I specifically was focusing on clustering So the idea of identifying trends not necessarily identifying for each building what they're doing it's a larger community know we want to see where the trends are with what is the high end low or mid user and then calibrate the building energy model and validate it's use to basically see the measured compared to the simulated and then create the scenarios heating cooling lighting equipment how can you change aspects of design so that it's better and create eco feedback loops where we have a dashboard that basically allows the users to know at this point in time what can I do to become more energy efficient so without going too much into the technical because of limitations of time we're basically understanding each building's energy use as profiles and then creating from these profiles these trends using that information criterion that is are telling us what are the high transmit or low specifically this is specifically for a 24 hour cycle for cooling imagine this apply to 8760 hours of the year based on the measured data and then we're using these scenarios of high middle low energy users to design occupancy patterns representation in simulation when people are in buildings how are they behaving and then getting energy outputs for cooling heating lighting and equipment heating really is not very relevant to to Austin in this case it's specifically cooling that's the biggest driver and so this is how the urban building energy model looks like it's a simulation for that community for the high end and low 3 representations designing some scenarios for how you can become more energy efficient for example deploying shading or changing set point so making set points higher for cooling for example and so for each of these we had these strands to compare to a baseline and one of the most effective ones was basically change your set point immediately you're going to be saving energy there's going to be an impact on conference and there's going to be some give give and take but you have to.

[00:26:26]
Understand the impact through that then we created these scenarios that are also combined so not just one scenario multiple scenarios coming together and seeing how we can communicate this so we designed at the board that Dashboard is in its early phases this project is funded by the National Science Foundation in its early stages the seed funding just ended we applied for more funding so if I have something to show you maybe you will get rewarded by the n.s.f. and we will tell you more about that in 40 years or maybe not but at this point this dashboard looks like that's where we have a comparison between President historical data so at this point in time you see what's happening before and what could happen in the future we have energy awareness icons that give feedback to users so it's a smiley face that tells you you're doing well in comparison to your community or you're not you should be doing better selectable actions based on the simulation so we have these tools that you can press on what if I deploy cheating Now what if I change what if I turned off lighting and finally we have to have this for heating cooling lighting and equipment to inform your decision much better so the whole idea is we're trying to communicate to people something about their decisions and about the uncertainty that they're going through and how can you become better so to gauge your understanding of this concept I want to give you an analogy of it so how many of you have seen the Avengers movies.

[00:27:50]
How many of you have seen the Avengers movie so we're like talking about the 3rd 3rd of the of the audience you're not not too much which is actually completely typical I'm more of those people who really care about that movie but 11 protagonist in the movie was Dr Strange for those of you who don't know Dr Strange he has a time stone and that time still allows him to see into the future so he sees 14000605 alternatives of how he they can defeat the division at the end and at this particular point they're going to decide how they're going to take the best path right and in the last movie I'm not going to spoil it no way but basically they come to the one way where they're.

[00:28:26]
Going to defeat. The villain and then this point it's really a good representation for how I want you to think about this and how I want the user to think about this I want you to have a time stored where you're actually going to be seeing your multiple futures and then making that one decision that is going to allow you to well defeat this case and it could be climate change and with that thank you so much.

[00:28:49]
Look forward to discussion and everybody. So this is a project that I had a chance and a pleasure to work with Chris. And so to give you a little bit of introduction about the project so we work with children. And they're working on these projects and they're working and designing and you Hospital for Children and basically their goal these 2 what they did was they created this large cardboard up of future hospital and they try to simulate.

[00:29:28]
Every all that actions and activities in the space and so it's $1.00 to $1.00 scale and it's basically 1100000 square feet of you know hostile mock up. And so. We don't mind so our goal was recall average with them and help them to collect data but I also analyze data the whole goal is off doing the small card was to.

[00:29:58]
You know bring people in and collect feedback for medical staff and medical personnel on their workforce their task their activities but also how the future hospital would basically you know help them or support them in doing those activities activities and tusks So part of this is really human factor and what we did was trying it was all focus on people on how to use this space and we use different source of their collection methods and we try to help them as human centered designers to make sure that we understand the whole process better and we support them with it and also.

[00:30:42]
Make sure that we basically introduce a new approach we call it centered and now thinks in designing future hospitals. So to give you a little more. Description or more understanding of why we call it human centered. Fix. In general I think in designing word. The reason these assumptions human centered design is really different from.

[00:31:14]
Say they time now thinks and here we argue that they are complementary metals and they're not income conflict and they can help solving when especially when it comes to complex issues and critical issues such as designing hospital that we can tolerate so there won't be any 0 per cent of errors would of would be suitable right so we can there shouldn't be any to any race if we mistake in designing.

[00:31:43]
The hospital right so in that case we can combine human centered design and of course with data and I'll fix to ensure that at the same time while we do these rigorous analysis of. People's interaction with this space so and so forth at the same time he steals human centric so if I want to summarize what I said so human centered design is usually.

[00:32:12]
It can be messy process. So I think they to and can make the process a little bit more straightforward and in Portugal our help help designers and researchers to bake more straightforward to make better decisions and make more straightforward decisions. So in order to basically achieve those goals we help them beat being and specifically within the law and allies and they are a different source of data so every was one of them and basically the goal was to.

[00:32:47]
People's feedback after running scenarios in the space after using this space. Question about the conflicts and problems in the space and then we had. These trigger was observational study and we try to analyze all these videos that we recorded then corded and we help them with that of course at the end we try to add.

[00:33:12]
Carly we are trying to triangulate on and for fun all the conflicts and basically make sense. That a duck Plus what is what are the problem in the space from the standpoint of people using the space so I pass it to you to create. The definition. Of what it says we.

[00:33:41]
Have a service that describes a human. In that human sentiment across multiple categories questions right person for. The kind of. A number of humans are crossing into the protesting areas. This morning. And then from there we worked we consolidated the questions the sentiment into categories such as committee communication collaboration the fish and sea medication.

[00:34:14]
And patient safety right and so we're looking out for this before and after comparison this is around one round to the improvements in the differences in sentiment across those dimensions of her discipline crisis here Richard or in a primary or to use rights of these are people that were brought in by the specific scenarios inside the area and they want to pull back and look at the testing area.

[00:34:40]
And see. You know click one zoom outright so testing here is going to have a mix of these people performing certain procedures that the snake use new natal intensive care right so if you have a different it's got a different wind of people and so we want to look at.

[00:35:00]
A stay score change across the testing period writes an effect we're trying to. Quantify the subjective human experience by display and by testing. Some of descriptors. So you can foresee. You know it gets up to the time. Of school as well as here you know which we brought to the you know.

[00:35:29]
The do. You know what it says for 2 for the diagnostic it whether it's which comes from the video that. Right so. I just wanted to get to show you an example that's one scenario again and I see that basically as you can see on the call 1st of all trying to fish same time we went on and sold this label and.

[00:35:56]
If you pay r.t. and you. Say and what we found also shown again and then and now the best thing we found that they're very accurate and function and I'm eating chocolate and comfort. Food to begin person with them and equipment and so that assures us that yes there is a problem so this scenario for example beastly YAGNI me it's.

[00:36:26]
A serious every version fighting test is designed so to pass it to you to continue. Through your core tool right so. Will we be true to the dark not. For. The. Interest in your interest in the troops. Or listen to Rush try to if we prove the subject experience for richer or ill there's a hardcore wishful.

[00:37:01]
Armory or do we care what the general we can do our best to do was for her to do this is to improve your performance based on those words you know. Through. Her the support of. The world the looser comments from the server. But with the whole translation they are the.

[00:37:36]
Subject back to design a relation we can. Push we wish to design to work with so that. I think it's a good time to any Let's have any. Question They pick a ship question from each of us and in each of please send a good senior editor going through the last.

[00:38:10]
You. Know. The. Most. Recent. Real threat especially with the partners that show up on the loop really the literature says the what the r.n.a. order is great that it is. Training simulations right so if you brought a position in that it's huge and it's usually position training right now right because you can put them in of your environment and have them practice this procedure and experience actually we.

[00:39:01]
I think that. You would get richer feedback and richer insights from your users if you can have them assume the same mental model that they're going to be running inside a real scenario in the end in a simulated environment is much better at that than of the argument.

[00:39:21]
So is your question. Yes really. Yes. So well thank you very much for the question and I actually think that most of what's happening at this very moment as in when we are building now we are attempting to cater to people's behaviors and to make it as energy efficient as we can't the unfortunate thing is that 50 percent of the built environment at least in the United States both commercial and residential was built before the year 1970 so most of the built environment is not going to be responding very well to the energy efficiency measures we need to be much more reactive to a climate that is changing drastically so one way is to attempt to bring the information the analysis of the data to inform the decision it's in their current environment Another way is to design the environment that had becomes energy efficient.

[00:40:53]
Doing both in tandem will not be harmful. Basically responding to how people can how we can inform them to teach becomes very important because not everyone is actually aware of how their impact is better and one way to do this might be competitiveness like knowing how other people are doing may be a driver it is really not an easy question to answer how we can make the built environment more energy efficient that's one approach to it basically but thank you.

[00:41:44]
Zack. No not Senator I agree with you on the cost is just one number that you see at the end but the ingredients of that and elements of that are really important I see movements as far as the. Democrat is ation of the data you know typically like many of these things comes from Europe but there is an international cost management.

[00:42:46]
Call international i.c.m. standards for how to report the costs and yeah a lot of owners adopt at. Baseline and benchmark I would say you know if they're large owners of capital projects like Georgia think decide to get a similar kind of data and since there are probably. So we in some 6 there is we have better data compared to the others that came from a structured for instance because all of that is for the most part the public we have better data and you can you really know what your company's if there is data on the past and projects.

[00:43:22]
But. Will they usually start back toward like everybody has a My understanding is they have a profit as far as that to get for their own profit and then they start back so the remains of their cost estimate knowing that they want to make this much money on that project and that profit margin changes as the.

[00:43:49]
As the climate changes you know competition and different type of projects but yeah you're right you're right you know the best sources of that are you know you the general recommendation if you are a general contractor which is might be 8090 percent you don't perform you have to get to your stops so for instance you guys own and they have to be called site so if you called subs they have the best of their challenges I can tell you the number one challenge is the estimation of their productivity production rate which is only at the crew level you know what would be I mean at the high level if I use average old war scary scenario of a scare scenario it's really I really don't know what that project looks like so I think bringing gone Dick the crew who would be potentially walking around the subs have the best best data and most accurate data.

[00:44:45]
Your supporters. Here. Want. You. Yes Yeah no it makes great sense so 1st of all I must actually preface this by saying that the mother is actually really not the typical it is very highly granular no one actually measures appliance level data the minute scale it's really too much but that said measured data exists as long as we are doing 2 activities mainly one is ensuring buildings and the 2nd is.

[00:46:14]
Paying for utilities so the utility companies have granular data that will very reluctantly and rarely share that is going to be the biggest hurdle right but I have seen cities in Europe like Lisbon in the Middle East like Riyadh in Saudi Arabia and in the United States like Boston invest in building an energy model to understand energy flows and identify for example for Boston heat sinks where.

[00:46:44]
We can know where combined heating and power would be most effective geographically for example so it really depends on a policy level question 1st how can we share that data privacy matters come into mind immediately and then when we are resolving that we can definitely use that data to get something out of it it's not going to be as granular or as effective when it comes to the Mahler case but it's definitely going to have some impact of change we need to change at the policy level 1st for sure thank you.

[00:47:29]
It. Was. My. Fault. Says. The. Mission. Yeah yeah that's your collecting information that can be used against you that's that's a no then reality is if you are a good project manager. Yeah you can use that to protect yourself but. It's a very difficult problem you know if you are you know now with the.

[00:48:49]
We talked about these are the digital infrastructure and also this connected software is all on the clout then the challenges whatever actions you have is so on the restored so you cannot say I didn't see that I didn't see that part of the model or I did not review that or thought I forgot that e-mail or something like that this already dear soul yeah the whole ramification of that you know I think it's something that you have to be seen I completely agree with you.

[00:49:31]
I would say at least for the for the time being no for maybe talking about next 5 years but one point to be half year we would like you know this is a research for what the we see opportunities out of college and the whole built environment domain investing in an Alex then an architect's coming out of a school or a city planner as part of their professional licensing and professional obligations is they need to have some good understanding of these basic understanding of how to use anality and therefore their their professional liability actually expanded to that so if you see the project is running through the trouble and you didn't say anything that they can hold you accountable I think at the definitely there is a legal way moving forward I know we're kind of running out of time I just give you one example about the traffic just Imation everybody's thing you know the traffic estimation is traffic traffic count and you know what would be the travel demand for the city is always subject to uncertainty but there is the recent case as all those one year ago.

[00:50:40]
Open which is a big u.k.. Based international design and planning firms they got sued for $1600000000.00 in Australia because their traffic model was biased and did not address that so and they they paid that so that's a lot of mean things for those who are just say no my job is just to give you a future scenarios where you know when you are entitled for if my revenue depends on your projections I want you you've come up with that and it's very expanded liability you touch on good point.

[00:51:18]
If you feel very sure of. All of. Course you have the. Price proportionate down here. And 1st. There's so many and. One. But sensible. Yes. Thank you.