00:08okay hello everyone my name is jeffrey
00:10olette and i'm an open bim consultant
00:12for the built asset industry
00:14i'm here today to host the industry
00:16perspectives on scan to bim panel uh
00:19discussion on behalf of the open design
00:20alliance i want to thank all of our
00:23panelists for joining us today and
00:25before we start this discussion i'd just
00:27like to go around the group and have
00:29everyone introduce themselves basically
00:31with your name your title and and your
00:33company so i'm start you off andrew sure
00:36so i'm andrew fox sketch partners um
00:38architected catch partners been here
00:40about eight years now
00:42great dan uh dan smilow director of
00:45process innovation at the walsh group
00:51manager with pencil phelps for about
00:53nine years now great hector
00:56hector camps here fight cube i
00:58specialize in building life cycle
00:59management and gustavo
01:01hi my name is gustavo pardo i'm the
01:03design technology manager for perkins
01:08great again thank you gentlemen for
01:10being here so we're here to talk about
01:12this idea of scan to bim where the the
01:14results of a lidar scan you know the use
01:17of laser imaging to scan the environment
01:20and generate a data set also known as a
01:24it's then converted into building
01:26information model geometry and
01:28usually it's some kind of proprietary
01:30platform format like an autodesk revit
01:33or a bentley systems microstation file
01:37um oda has seen the growing demand for
01:40such a workflow and and you know that's
01:42what we're really here to talk to you
01:45finding out what those workflows are
01:46your experiences in them what you um
01:49what you think you need what you'd like
01:50to see happen in the industry
01:53and and they're really interested in
01:54helping the the bim software developers
01:57create more powerful tools so they can
02:00enable faster and more accurate scan to
02:04um you know as i said i think you know
02:06there's already software out there we've
02:07seen this sort of come about especially
02:09in the last uh 10 years we've seen a
02:12rapid development in maturing in the
02:15but it still seems like we're not there
02:17yet and um so i'd like to really you
02:20know start this conversation giving an
02:22overview of sort of where scanda bim
02:28my first question is really
02:30you know looking at your practices in
02:35how often is your company seeing
02:36projects where there is a need to create
02:39models from these scans and the point
02:43and think about this in three ways where
02:46how does this frequency compare to five
02:48years ago and do you think it's going to
02:51increase in the future
02:55so let's kick this off with andrew
03:00within our office i've worked on
03:02probably 10 of these projects now that
03:04have done use some form of scanning
03:06converting into something whether it's
03:08gone into bim with probably about half
03:10of them some of the early ones were
03:12scanned to flat cad drawings
03:15and there's been a couple projects where
03:16we've just received the point cloud and
03:18it hasn't really gone beyond that
03:25um within our our firm the section that
03:28i'm working on is largely renovations
03:30and preservation and restoration
03:34there is definitely an increase in
03:37demand for the use of this it's to the
03:39point that basically every project now
03:41in the last couple years has had some
03:43form of scanning whether it's
03:48the last project was a substation in the
03:53with complex floor and ceiling and wall
03:57and maybe at its largest scale we
04:00scanned an entire um auditorium space
04:04uh six floors and um and for a
04:08uh for many future projects but for the
04:10first was for a seating project where we
04:12really needed to have a great
04:15understanding of this curving bowl of
04:17seeding that then needed to be modified
04:20for for ada upgrades and
04:23all the slopes needed to be very precise
04:27to answer your questions um
04:29there's there's definitely demand and i
04:31think it is as long as we continue to do
04:33renovation projects it will definitely
04:36be an increased demand too
04:41agree 100 with andrew you know majority
04:43of the time that we're leveraging this
04:45technology there's kind of
04:47three reasons one the owner's driving
04:49that they they require it in some form
04:51of specification and we're providing
04:53that whether that's a as built turnover
04:56or something along keeping up with as
04:58built throughout the the building cycle
05:00of the project but where we see the
05:02greatest value is in those retrofits
05:04those renovations um
05:06walsh has three different verticals it's
05:08our building group civil group and our
05:11water time and time again when you're
05:13massive upgrades how are you
05:15retrofitting this existing facility with
05:18new pipings new filtration systems new
05:22really having a granular view on how
05:25this is going to integrate into the
05:26current process and procedures is very
05:28difficult and you know can go out there
05:30and take a tape measure and measure it
05:32but you know that really doesn't provide
05:35great hindsight when you're throwing
05:37these models together
05:39as well as how you're going to get this
05:41information or this equipment into these
05:43spaces um other places that we're seeing
05:45in distribution centers everyone's kind
05:48trying to retool themselves or upgrade
05:50their distribution centers that are
05:52already packed full of distribution
05:54equipment so how can you get the best
05:56fit or best space layout in that and
05:59laser scanning has been great in that
06:04yeah for us it's it's very similar um
06:08deployed point cloud technology for last
06:13via either laser scanning directly or
06:17um increasingly with with drones to
06:19capture the larger type sites and
06:21i would agree that um that this process
06:25it's it takes a lot of time so really
06:28going into the project it's it's that
06:31typically it depends on our trade
06:34whether or not they have that capability
06:36to deal with the put cloud directly or
06:40we need to execute that that conversion
06:44we see it a lot in our aviation airport
06:47projects large spaces where
06:50we're getting a lot of benefit from that
06:52your your laser scanning can capture
06:55above ceiling content which is extremely
06:57valuable during coordination
06:59and then just the ability to to convert
07:02that to a 3d we typically shoot for
07:06autodesk based programs but it's
07:08definitely dependent on
07:10on the owner but the ability to work
07:12in that environment as opposed to
07:15how heavy these point clouds are
07:17it's definitely an advantage that we
07:20and it's something that also feeds a lot
07:22of the other technologies such as
07:25augmented reality devices with the
07:29quite well with a very light model
07:31that's been optimized
07:34and so it's it's just that mix where we
07:36see the opportunity we'll we'll
07:37definitely go for that
07:39and it also just depends on on the
07:45yeah and so these are all really good
07:48i tend to look at it from the building
07:50life cycle perspective so i always like
07:55why do people want to do the scan in the
07:58front end the design construction
08:00process you have the as built condition
08:05front end laser scanning work that we've
08:07done in the past has has been around
08:10capturing the ass built condition uh for
08:12example we did mammy international
08:15airport which had an extremely large
08:18footprint you know it's an airport of
08:22where nobody really trusts the as-built
08:24documents because they go i need 10 20
08:2630 generations you know they're like
08:30you got you want the 1975 version or you
08:32want the 1982 or do you want the 1994
08:37and they recommend you look at all of
08:38them because they all have different
08:40pieces of information on them but none
08:42of them have the truth
08:45because you can't trust the as built
08:47documents that we're inheriting from
08:50the history of that building
08:52and people want the confidence to truly
08:54know what's out there
08:55we've been scanning to document the as
08:59another interesting reason why we've
09:02you know of course you can take the
09:04professionals and bring them to the
09:05field but it's much more interesting to
09:07bring the feel to the professional
09:09especially when you have remote
09:10distributed teams that could be anywhere
09:12in the world uh the people who are for
09:14example in this case
09:16that were working on the um
09:18baggage handling system for mia
09:20they were out of michigan
09:22so they couldn't really fly to michigan
09:24all the time to see what's actually out
09:25there so this was a very effective way
09:28to bring the airport to the team
09:30that was off-site and remote and needed
09:32remote access to see what's actually out
09:35there so that's uh that's part of the
09:37reason we've been doing this
09:39the second reason we do this is to do
09:42field verification so
09:44we can really get involved with
09:46constructing from the model we try to
09:48drive a construction driven process from
09:52and we built you know there's bim and
09:54then you also have vdc right virtual
09:57design construction which gets into more
09:59of the construction fabrication side of
10:02and since we're driving construction
10:06we need a high fidelity model we need
10:08confidence we need fit form we need to
10:10make sure that those structures those
10:12baggage handling systems whatever they
10:14happen to be are actually going to fit
10:17correctly the first time and we rely on
10:19the model to be able to perform that and
10:25you make my outline difficult you really
10:30in such a great way so for for us in
10:34parkinses man you know we're large firms
10:36we have large projects um
10:39existing conditions are fundamental we
10:41we really need to get existing
10:42conditions uh there is not enough
10:45interns to go to a place and you know
10:49and um and i would say
10:51we probably can divide this thing in
10:53three areas one is coordination you can
10:55say efficiency that goes with
10:57productivity and then precision so
11:00if we're talking about
11:02an airport or if we're talking about a
11:06it's the same you know we need that
11:07precision uh another thing that hector
11:09mentioned that that i think is
11:13on critical is that so many times we go
11:15to a place we get the drawings you will
11:17get the really old autocad drawing so
11:20you really get the all uh bean model and
11:23as soon as you go to the side you're
11:26what happened with that window or what
11:27is that wall and and if we think about
11:31going back to that efficiency and
11:35by the time that your team goes back to
11:37this to their place or by the time that
11:39you have your first condemnation
11:40uh call and you realize that the wall or
11:44the window is not there you probably pay
11:46for half of that uh 3d scan so if we
11:50think about that you know in that call
11:51you're going to have probably
11:54high-profile employees that does a lot
11:58of money plastic time
11:59um something for us that that is
12:01important is the scale and we have some
12:05we have towers that we want to retrofit
12:07we're definitely going to do that but
12:09there is also a niche in in a a practice
12:12area that we call workplace
12:14and and there is so many smaller spaces
12:16and uh and we see the drones as gene
12:18mentioned uh we see the point clouds we
12:21see all type of 3ds kind but there is no
12:26that small ditch when you have to 3d
12:28scan something you don't want to send
12:32but you need to capture the space so
12:34there is a lot of opportunities in the
12:37in the field we have seen increasingly
12:40um i would say probably
12:42a hundred percent of the projects that
12:44we have that are existing conditions
12:46they're doing 3d scans it is not you
12:49it's a pretty easy decision it's a given
12:54yeah yeah okay excellent
12:56so you've all mentioned the types of
12:58projects that you use that that
12:59typically are doing scanda bim
13:02maybe you can uh help also describe what
13:06is your process i mean are you
13:08are you going out and doing the laser
13:10scanning yourself with your own units
13:12are you hiring somebody third party to
13:14do it are you then getting the point
13:15cloud data and and translating that
13:17yourself or is that also a third party
13:20what you know what kind of things are
13:22you relying on i'm going to skip over to
13:27probably every variation and then some
13:29of what you just mentioned um depending
13:31on the complexity of the project
13:34the availability of our scanners you
13:36know that's that's always a
13:38difficult thing to approach we have
13:39multiple scanners within our
13:42um you know we have a retrofit project
13:44comes up you know that's that's just
13:45like a full-time fte that's sitting
13:47there and that's being leveraged on that
13:50one of the banes of our existence is
13:52when we hear we need to perform laser
13:54scanning on that project and the first
13:56question that comes to mind are we
13:58creating the model oh no the architect's
14:00creating the model no problem we'll go
14:02scan that that's we're completely fine
14:03with scanning that project because
14:05that's the easy part scanning and
14:07registering those points no problem uh
14:10the complexity of really developing that
14:12model that's time consuming and
14:14unfortunately these days in the industry
14:16it's you're closing up one project
14:18starting up another project
14:19simultaneously and bidding a project in
14:21between the two of those so um it's very
14:23difficult to kind of stop do all those
14:26iterations and go through it we will we
14:28have and we've done that before where
14:30we're scanning and we're creating those
14:31models um traditionally we're hoping
14:34that the architect is part of their
14:35scope or um we will bring on um
14:39consultants or contractors in that we
14:41have trusted relationships with and
14:43they'll create that that model from
14:45those scans as well so
14:47we probably use every variation of it
14:49and you know we really determine
14:51what's the best utilization and time of
14:53this model knowing that this project is
14:55going to be a very complex retrofit well
14:57we should be all hands on deck on that
14:59project if it's something that it's
15:00going to be continuous scanning and it's
15:03not necessarily for a model it's more
15:04for as built we might have a third party
15:07come out and do those scans every couple
15:10it's not kind of part of the build-out
15:12workflow it's more of that close-out fm
15:15deliverable at the end of the day great
15:17how about you hector
15:20well he i want to pick up where daniel
15:22left off so that fm deliverables are
15:24really interesting um
15:29our customers are becoming savvier as
15:32and ownership is saying hey wait a
15:35we're really interested in this scan
15:37beyond the construction
15:40if we ever need to go back there we want
15:42to see the scans we want to be able to
15:44kind of see the bones of the building at
15:48different milestones and we want to be
15:50able to kind of turn back the wheels of
15:52time and be able to see the building in
15:54his different historical historical
15:56points so i think this is a really um
16:00a really big point a lot of people are
16:01starting to associate data and
16:04information with the point clouds as
16:07so i think that's um
16:09kind of an emerging thing that we're
16:11starting a trend that we're kind of
16:13paying attention to and we're seeing
16:16tracking the historical development of
16:18the project through the scan is becoming
16:21a a record document of source that
16:24ownership is now starting to expect and
16:27wants to associate with the model and
16:29the documentation so we're seeing a lot
16:31of that the other thing that we do
16:35we do a lot of field verification um as
16:38daniel also mentioned
16:40when we have the construction schedule
16:43and those milestones we generally look
16:46at the schedule we try to figure out
16:48when it would be a smart time to go out
16:50there and do some field verification in
16:52essence to make sure that the building
16:53is going on track and they were getting
16:55it built right the first time and to get
16:57some foreshadowing of what's going to
16:59happen if we continue this trend and
17:02we're doing this with the scan so we'll
17:04you know the mep goes up there the
17:06mechanical systems go up there we want
17:07to scan they pour the decks we want to
17:09scan so we're basically following the
17:12construction schedule and scanning at
17:13different milestones so are you doing
17:15that with your own house equipment or
17:17are you using a third party to do that
17:22we lease our equipment from fireworks or
17:25from a third party so we do our own
17:27scanning and it and we've been um
17:31i'm using more of the automation where
17:34we're trying to use more of the auto
17:36registration capabilities which cost us
17:39cut some time for us
17:41but if we if but if we're in a snag and
17:44we have a very complex scan to put
17:46together we may rely on a third party to
17:48help us put the scans together or
17:51basically what's called registration how
17:54so from the architect's standpoint um
17:57we've seen a couple different variations
17:59we've never done our own scanning we're
18:00usually contracting out it's either with
18:03the general contractor
18:05um or it is coming from an outsource
18:11and then from that point once we've done
18:12the scan we have both modeled in-house
18:16and we've also outsourced the model or
18:18the general contractor has provided the
18:20model from their scan it's our
18:22preference not to do the um
18:26make the conversion ourselves because it
18:28just isn't time and cost effective to do
18:32uh but we've done it it's
18:34it can be a couple week process for very
18:36complex things if not a month-long
18:39process for much more complex things um
18:51recent projects as i said i've moved
18:54into we've moved into this outsourcing
18:56process there's a couple firms in the
18:58city that we work with regularly now
19:00uh who will do the scan
19:02create the bim model we usually like to
19:04see the scan and the bim model even
19:06though the scan file is usually large
19:08and sometimes overpowering for our our
19:12there's a lot to be gained from be able
19:14to overlay both them over each other as
19:16we move into the project
19:18and being able to see
19:19you know what areas do we need more
19:22uh what areas you know they're not
19:25always 100 accurate where where can we
19:27catch really important places where the
19:30scan and the model might vary even
19:33because the building might vary a little
19:34bit like it slopes from one side to the
19:36other and it couldn't be accurately
19:38gathered within a bim model
19:43how about yugoslavia
19:45um well again it depends on the scale um
19:47if it's something that is is you know
19:50small enough we just go out there and
19:52just take some measurements uh most of
19:54the we don't have the equipment and
19:56everything that we do we outsource we
19:59we're great designers we we build
20:01beautiful buildings but
20:033d scanning is now an area that that we
20:07that we want to expand our time
20:09and uh the same thing that i do mention
20:11you know for us is uh
20:13it's better if we just get
20:17um already in in beam and which is
20:22into our models um something that that i
20:25want to mention is that even when we get
20:27the 3d scanner and this is a project we
20:29have in new york we got a 3d scan from
20:34we do some some fuel uh measurements and
20:38really quickly we realized that even the
20:403d scanning from the client was
20:42different to existing conditions so i
20:48and i see this this is it should happen
20:50in the future somehow facility
20:52management the whole process of
20:55us building those beautiful 3d models
20:57having the information that hector was
20:59mentioning that we seeing all these qr
21:01codes in the rooms when you can start
21:03seeing what is inside that room
21:05but it should go farther you know it
21:07should we should have we should find a
21:09way to track how our buildings are uh
21:12evolving through decades
21:15yeah i think jim you mentioned a little
21:17bit but are you doing this fully
21:19in-house or are you kind of
21:21as dan mentioned is you're utilizing
21:23whatever you can whenever you can for
21:31the complexity of the process
21:34a laser scanning piece
21:36that's pretty much done 100 in-house
21:41laser scanning has become
21:43a pretty important process to our
21:45construction in general just for
21:48these these renovation type spaces to
21:50get an accurate as built but also
21:53to check the quality of slab placements
21:56and document all that during the
21:57construction process but a big
22:00a big point is it's something we've
22:02toyed with as far as the the model
22:05it's something that we used to
22:08handle all of that in-house of the
22:10software years ago has been developing
22:14automatic feature extraction
22:17but what we found is
22:19particularly with piping type systems
22:22the software will create those elements
22:24but then you spend just as much time
22:27seeing making sure everything fits
22:30we weren't really seeing the the
22:32efficiencies to make that process go and
22:34that's that's kind of why we're here
22:36talking about this the
22:38industry's been chasing this magic
22:42allows us to capture more data than
22:44we've ever had on the construction
22:46site and then with drones as well
22:49but the ability to efficiently convert
22:51that to a model is is something that
22:54really holds a lot of people back
23:00person on the site has a super computer
23:02that can handle that point cloud
23:04and so for us it's it's a mix just like
23:06everyone else for for smaller type
23:10we'll handle that conversion but for
23:13projects such as airport terminals or
23:16baggage handling systems
23:18we have a lot of third-party
23:20partners that that we're comfortable
23:23um but even so that's that's
23:26working with different groups and
23:28getting that process dialed in between
23:30uh between us and the partners we use so
23:33when i do a little follow-up if you
23:34don't mind so what would be the role of
23:37automation in 3d scanning because we all
23:40face the same challenge when we go and
23:44we get a lot of information that
23:46we probably don't need so for example um
23:50i get a 3d scan i come with a bunch of
23:52piping that has they're not really
23:56let's say this copper work that we're
23:57doing i wonder if in the future some
24:00kind of algorithm is going to start
24:03what is piping what is a wall you know
24:06it should be something that starts
24:08filtering those things we we have seen
24:10some automation with you understand what
24:13the what type of geometry is there and
24:14say okay this could be a wall and it
24:17and it throws a wall there but it should
24:19be something that it would it moves
24:21further i don't know where where the
24:23technology is going to what direction is
24:25going to go but for now i'm just looking
24:29it is a window has transparency so it
24:32will create a material
24:34i don't know if anybody has any
24:35experience to have seen any out there
24:37that start moving towards that direction
24:41i see some work being done with lidars
24:45the lighter scan can tell what's a car
24:47what's a building with a tree what's the
24:50ground and you can filter out the lidar
24:53scan well it can take out the vegetation
24:55from the from the scan and just give you
24:58the floor condition for example so i
25:00would imagine in the near future we're
25:02going to be able to see something
25:03similar to what you're talking about
25:05gustavo um with connecting it to
25:10again i've seen that the larger scale
25:12when we're talking about scans of cities
25:15you know what dams and this sort of
25:19i haven't seen it all the way down to
25:20the granular level of a building yet i
25:22mean i have seen some experimentation
25:26but it's it's sort of in the beta stages
25:29what i've seen so far
25:31so it sounds like everybody sort of
25:33agrees that some of the biggest problems
25:35that you face so far is a
25:38you get a lot of information you get a
25:39lot of data from these point clouds and
25:41sometimes it's too much it's it's either
25:45too much in that the data set is so
25:47large it's unconsumable by hardware by
25:49the systems that you have right or it's
25:52just a lot of noise i mean there's a
25:54hell of a lot of noise right and you
25:56know in these scans they can vary their
25:59their uh precision input to
26:0210 is it 10 points within a square inch
26:07um and can go larger than that so you
26:09know of course the larger the area that
26:11you're scanning the more points you have
26:12and that can be very dense
26:14um sounds like cost and the time
26:17associated with a making the scans and b
26:20making those conversions is a very big
26:23uh you know obviously not every
26:26not every type of stakeholder you know
26:28if you're an architect and you're
26:31that may not be feasible within your
26:33firm or you know sounds like on the
26:36contracting side you figure out ways to
26:38make it work because
26:40that is a part of your process that is a
26:42fundamental part of what you do and
26:44being able to do the scanning but maybe
26:46conversion isn't always necessary or if
26:48you could get conversion on top of that
26:51as an easy gain then it would be much
26:53easier to to justify
26:57okay so you know i think that does a
26:59pretty good job of laying the groundwork
27:04you know so so oda is is
27:07you know in this this realm of trying to
27:09provide sort of these core technologies
27:12that fill these gaps right if you if you
27:14look at what oda has done so far far
27:16they've provided a lot of
27:18interoperability toolkits for the
27:22you know getting dwgs converted from
27:25distant different systems now they have
27:27an ifc toolkit they have a revit tool
27:30kit they have a dgn part of their
27:32toolkit and the whole idea is to enable
27:35this information to travel between
27:38so this approach essentially from oda's
27:41say well could we do the same thing
27:43in this scan to bim you know range and
27:47you know in a tackle that says well
27:49let's provide the tools and the
27:52platforms for anybody and everybody to
27:56you know and and as a cost to the to
28:00um but what do you think about that
28:02approach i mean does it does that seem
28:03to make sense to you rather than sort of
28:06you know each of the individual vendors
28:08and sectors to just sort of figure it
28:14well yeah i think there's a need so it's
28:16interesting because um one part of this
28:18is how do you consume the scan data
28:21one of the issues that we've always had
28:23of course you know when you talk about
28:25uh lasers kind of an airport you can
28:27imagine how difficult this is to consume
28:30in every sense hardware just
28:33transferring the information making it
28:35accessible making it available bringing
28:37it into other systems so we obviously
28:40need more optimized ways
28:42of being able to make that data
28:44available in a wide range of platforms
28:47in a wide range of environments
28:52you know tablet computer all the way up
28:54to your smartphone you know how do we
28:56get a laser scan on a smartphone and
29:01so we've we're working on finding more
29:04optimized ways of delivering this data
29:08and making it consumable
29:10especially over the web and on mobile
29:13and remember one of the big reasons
29:15people do this is called metrology right
29:17it's all about me you know one use case
29:20is just being able to measure
29:22and pulling information off of that the
29:24scanned data and being able to do that
29:30so there's of course you also have uh
29:32being able to geo-reference the scan to
29:35when you're standing in a in a location
29:38and you want to take your smartphone and
29:40kind of like use augmented reality to
29:42kind of see what's in the ceiling and be
29:45able to pull in the scanned data this
29:47sort of thing and tying it back to a geo
29:51and again being able to tie into mobile
29:54it's all about making it consumable and
29:56making it extensible throughout the
29:58design construction process
30:01and anything you guys can do in that
30:02direction would be fantastic
30:07and i i would definitely say the the
30:08burden of entry that's that's the
30:12toughest thing in in our industry
30:13regardless of whatever tool or device or
30:17especially in this one where you're
30:19talking about when you enter this space
30:21of laser scanning you can get third
30:23parties you can lease them but if you
30:25truly want to have a holistic program
30:27in-house buying the equipment it's a
30:29six-figure entry cost
30:32so to get to that point just to
30:34understand something that's a pretty
30:38once you get there you start scanning
30:42this data set and now okay you have to
30:45one of dozen different solutions
30:47okay i have this beautiful wire mesh now
30:49what the heck do i do with this
30:51well now you got to take it into another
30:53program and then you have to edit it and
30:54you have to build a model you know so
30:56it's how do we how do we simplify this
30:58even more how do we make select
31:00selectable meshes how do we make that
31:02where we're working is more in the mesh
31:04and less on modeling it that that
31:06active participant if you will for the
31:11you're seeing movement on the laser
31:13scanning side of things you know it's
31:14everyone's got the newest latest and
31:16greatest cell phone or ipad and they now
31:19all of a sudden have laser scans on them
31:21nobody knew what the heck they were
31:22going to do with it you know it was for
31:23the ar so you can have a little dancing
31:25rhino on your desk but then people in
31:28our industry started being like wait a
31:31work with an interior designer i could
31:33scan my apartment send it off to
31:34interior designer and get a beautiful
31:36decoration of my apartment so
31:39if we could do that in our personal
31:40level why why aren't we trying to
31:42holistically approach this in our
31:44industry to make it easier if i can go
31:45out there with an ipad all my sub or all
31:47my formative ipads out there let's start
31:49scanning this let's start tying this
31:51into different third-party solutions uh
31:53procure not procurement uh production
31:55tracking things like that so
31:58i think the sky's the limit when it
32:00comes to developing solutions for this
32:03because it's an underutilized and the
32:05burden of entry is so high that not
32:07enough people are playing here yet so
32:11majorly under underutilized uh principle
32:14in our industry and i i think it's it's
32:16worthwhile and it's worth the investment
32:17to start understanding what it could do
32:19for you and your projects
32:20yeah and i think you know i think this
32:23idea that the oda approach to
32:27you know one of their partners is
32:29building smart international and
32:31you know the building smart
32:32international philosophy is well
32:35one person can't solve at all so why
32:37don't we all work together to figure out
32:41for everybody for most the time you know
32:42figure out that 80 20 right and so that
32:45you know you're getting
32:47everything solved 80 of the time and
32:50that 20 percent outlier will figure out
32:53eventually and it may be a you know
32:56but working together to figure it out
32:59and then sort of harmonizing
33:01the the workflows and harmonizing the
33:06you know a larger spectrum of
33:08stakeholders and processes and things
33:10i think has a lot more benefits to the
33:13industry than you know just saying well
33:16you know if i've got 50 different
33:17companies are going to try to do this
33:19and they come up with 50 different
33:21now all of you in the marketplace have
33:24to figure out well which one of us 50 is
33:27and it's like well what if
33:29none of the 50 are good then what do you
33:33right and how do you get them to work
33:34together and say well maybe these five
33:37are close or these 15 or 25 are close
33:39now can you all work together to figure
33:42out what's the best out of that
33:44so i think jim you know dan pointed out
33:46some interesting and so did hector about
33:51this evolves then into the future
33:53workflows and the potential you know
33:55from their workflows
34:12makes it very simple to view just your
34:13data even on the capture side with
34:15drones and then with
34:17we've been playing a lot with robotics
34:19with spot um you have ways to automate
34:22that capture process
34:24makes that whole piece really efficient
34:28one of those big gaps in the market is
34:32on the software side um just the ability
34:36take this massive amount of data and
34:40a much more efficient model that you can
34:42work with that you can share with the
34:44that you can consume that down the road
34:47this feeds your hair process this feeds
34:52the ability to to perform inspections
34:55with point cloud against the model
34:59investment that takes a lot of time on
35:01projects for us but for huge projects we
35:06preparing this model so that's
35:08to your point i i think that's
35:10definitely somewhere where the industry
35:12as a whole would definitely benefit
35:14there are a lot of different groups out
35:16there that have different tools and
35:19um i'm sure like most of you we we have
35:22to be pretty agnostic we have to use a
35:23lot of different tools
35:26there's a lot of different conversions
35:28um to get what you need each has their
35:31strengths that's their weaknesses
35:33and we're constantly going through that
35:36on the scanning and software side just
35:38okay where's the industry app who's
35:42new technologies new potential to it but
35:46bring all those collective ideas
35:48together in a simple
35:50efficient solution i i think is
35:54an amazing challenge but definitely
35:56something that would benefit the whole
35:58industry i would follow up with
35:59what you're saying jim said and i see
36:01the same thing so one would be the part
36:03of visualization that gives access to
36:05clients and everybody in the team to
36:07really quickly understand the space or
36:10even take a simple measurement you know
36:12now to be able to access that data it is
36:14pretty heavy and it is hard to to
36:17anybody in the team to
36:18just open it and take some quick
36:21uh and the other thing would be
36:22translation so how you take all that
36:24point point cloud and you make it
36:25accessible to whatever platform you're
36:27going to use it we cannot restrict our
36:30our users or whoever is is working with
36:33this data to only work in one specific
36:37software so we cannot say well you if
36:40you use this tool to translate the data
36:42and now you need to use auto revit you
36:45know or you need to use um archicad so
36:49it needs to be something called like
36:50ifcs you know that is is it can work
36:53across any platform and you can open
36:55this this file in sketchup if you want a
36:58rhino and you can get the same precision
37:01but it's all about being efficient
37:04so dan mentioned something interesting
37:07and kind of a question is
37:10depending on what your use case is
37:14that conversion of scan to bim might
37:17have different requirements right so
37:19dan was saying you know for us i would
37:22love to use a mesh right
37:24and maybe in some cases just the point
37:27clouds are enough in some cases
37:29a simple point cloud to mesh conversion
37:32enough in some cases you go that step
37:34further what you say well no i want
37:37constructive solace geometry that's
37:39typical in a bim system
37:43gives me parametric then parametric
37:46control over all of these features
37:48i mean so so it sounds like there's also
37:51a need for having that
37:53that user control over
37:55how far to take a conversion or what
37:58direction to take your version not just
37:59the fact that you have to convert it
38:04any other thoughts on that
38:08well on one side you have the software
38:10generating uh the geometry
38:13but that's in the world of bim is a
38:14little bit more complicated because
38:16you know you can make something that's a
38:18cylinder look like a pipe but then then
38:21you have the building object which is a
38:25right and has more attributes connected
38:27to it you know is it is this an electric
38:29pipe is it a gas pipe what kind of
38:31conduit are we looking at so
38:34it goes a little bit further to just
38:36generate the geometry you have to
38:38generate the building object for that
38:40geometry that's appropriate for its use
38:42and function of what it is
38:44um so and i don't know the software's
38:47smart enough to tell that's an
38:48electrical conduit versus this is a fire
38:51protection pipe here so i don't know if
38:53we're there yet in taking it to that
38:57would be cool i mean artificial
38:59intelligence might be able to get us
39:02but just based on geometry alone i don't
39:05know if we can go the full mile with it
39:09with the building object situation
39:13go ahead and i was going to say there's
39:14probably a step before that where
39:17where where maybe the the future and not
39:19quite attainable yet is is being able to
39:24um point cloud and come up with an
39:26actual you know bim construction
39:30type level of detail
39:33in between there is kind of perfecting
39:34the the um point cloud to mesh system
39:38which could be useful even going back to
39:40our recent project of trying to
39:44decipher between seats and floor in a
39:49theater floor and trying to find little
39:51bits of floor that managed to be scanned
39:53between the seats and then create a a
39:57mesh out of that which was something
39:58that we didn't have the ability to it
40:00may exist out there but we certainly
40:02didn't have the ability to automate that
40:04and it became just a process of cutting
40:05a ton of sections and trying to figure
40:07out where the floor was among among all
40:12also to add to some of the other things
40:14that have been said here
40:15we have also been seeing clients now
40:19say point cloud data
40:21and short of me sharing screen with them
40:24there isn't the there are limited number
40:26of ways to actually send them say 300
40:28gigabytes worth of information that they
40:31need a software to to do anything with
40:36anyone else want to add
40:40okay so so one last question i have and
40:43some of you have already touched on this
40:45is sort of the things of looking into
40:47the future and how to use them i think
40:48hector had brought up a great example of
40:50you know using like ar with phones
40:54and we see more and more now you know vr
40:56trying to incorporate the the models
40:59um i've seen examples now of people
41:04those scans and your view and the models
41:07you know using an ar perspective
41:09um you know these are you know small
41:12startups or you know they're just trying
41:14to push it out there but i think a large
41:16part of it then comes back to well what
41:18is this what does this mean for
41:20your services to your customers as well
41:23as what happens to you internally you
41:25know if somebody had that ability to
41:28you know incorporate
41:30ar into it with these with the scan to
41:33bim and and then being able to display
41:36this stuff does that does that sort of
41:38change how you think about the
41:40information and then how you use the
41:49yeah my gut instinct on something like
41:51this i think um we're going to see this
41:54information become more extensible and
41:57like you mentioned jeffrey augmented
41:59reality and serving that information up
42:02on whether that's google glasses or you
42:05know what an iphone or technology
42:08iphone or a shield augmented reality
42:10shield in front of your helmet whatever
42:14as you know everything is going to cloud
42:17and point clouds are going to the cloud
42:20as well and if we could run those on
42:23servers they're running in the cloud and
42:27mobile device a lightweight mobile
42:29device whether it's your smartphone or a
42:31pair of smart glasses whatever the tech
42:33is in the near future
42:35i think there's going to be a market for
42:38consuming this type of data after the
42:40fact into field operations into
42:42facilities management into you know life
42:47or even even you know
42:49first responders going into a burning
42:51building scenario where they can know a
42:54little bit what's in the walls kind of
42:56add the x-ray layer vision
42:58um that has come from the scan data or a
43:01combination scan data and bin model
43:03so i think there's a feature for serving
43:05that up i'm not sure exactly who sir who
43:08who delivers this as a service but
43:11i think there's a future for uh
43:13consuming this data downstream and maybe
43:16that the building department
43:18now has a requirement to consume this
43:20data and make it publicly available i
43:23but i think i think there's definitely a
43:25feature for that okay
43:26that will get to a point that so what's
43:31maybe what is here is just
43:33the developing a universal format for
43:36point cloud data that can be consumed in
43:40different ways so then you can get okay
43:42the fire department has all the
43:44equipment but well that's not the right
43:46format so it should be some kind of like
43:48ifcs again you know some kind of
43:50international format that
43:52anybody can have access to the data
43:56i think going back to the original
43:57question i mean it's i think the sky's
43:59the limit on what you can and cannot do
44:04we're talking somewhat hypotheticals
44:07um on the most granular level what's in
44:11so what how is this going to benefit me
44:12and my workflows what are the benefits
44:15improving the constructability and
44:17improving construction telling that
44:19story um you know we don't just
44:21willy-nilly do things because it's cool
44:23and neat there's a purpose behind it um
44:25so what is that purpose and how can we
44:27get to that end result and uh just like
44:29the previous question on persona based
44:31stuff you know mentioned fire
44:33departments everyone watches ncis and
44:35this and that and the fire department
44:37shows up and they pull up the 3d model
44:39of the building and this and that well
44:41we all know that that's not a real thing
44:45what if you know what if we can get
44:49interesting discussions on the apple ar
44:51tags the the new tags that they have out
44:53there it's built on their ar kit so is
44:56apple secretly mapping
44:57our environments without us knowing it
44:59so you know how how do you get into this
45:02realm of uh real-time city diagnostics
45:05you get into digital twin a little bit
45:07you get into constructability reviews
45:09you know if i can really do that if i
45:11can pull the building up using google
45:14tear it apart throw in the 3d mod or
45:16throw in the laser scan
45:18pull different segments off add a whole
45:20nother floor you know you get into this
45:21really hypothetical world of these
45:25but it it's going to take that burden of
45:30how we can start to leverage this
45:32technology even more and you know what
45:34are we the the six of us on this call
45:35what are we not even thinking about that
45:37we're under utilizing in this realm and
45:39how can this truly evolve as this is now
45:43the replacement of the slide ruler that
45:45is now the cell phone in everyone's
45:46pockets because back in the day when we
45:48were taught math yeah you're never going
45:50to have a calculator on you all the time
45:52well and we know that that's not the
45:53case we all have calculators on this on
45:55all the time so in the future now that
45:57we have laser scanners on us all the
45:58time how can we start to evolve these
46:01processes and procedures to benefit
46:03ourselves and everyone in the entire
46:07hey i'll offer one up
46:13biggest banes on a construction pro
46:18the lag time of inspections to occur
46:22between stages right
46:26what if the inspectors didn't have to be
46:29there on site in order to get it done
46:31right what if you didn't have to worry
46:33about that kind of scheduling instead
46:35inspectors could use
46:38those laser scans to do the inspection
46:41and then sign off on them too based on
46:43that or make their notations and
46:44everything that they would do in person
46:48instead of say okay well uh by the way
46:51you know we're going to have a robotic
46:54at the end of every work day or at the
46:56end of every work week you know
47:01you know x date and you know and then do
47:03an approval on that what i mean to me
47:05that sounds like that would be a win-win
47:09i think that's the separate panel that's
47:12lawyers perspective so we'll need to
47:14come on to that one so
47:16well yeah and and i think even even with
47:19the scan to bim and and i think a couple
47:21of you touched on the accuracy of the
47:23the of the results right is
47:26what does that accuracy mean you know
47:28accuracy means everything depending on
47:30what scale that you're working at and
47:32then what is the ultimate purpose or
47:35goal of what it is that you're scanning
47:38if you are trying to make sure that you
47:42you know all of the walls in a building
47:44that may have shifted over the years and
47:47now you find that you have a
47:50a uh non-conforming condition
47:54you know that begins to pop up these
47:56ideas of oh well you know who's
47:58responsible for that where does the
48:01you know these things i and you know
48:04of course we all know in our industry
48:06you know liability and and all that is
48:10you know one of the biggest concerns
48:11everybody has because that equals money
48:13that you know risk and that liability
48:16you know can bring down entire projects
48:23always one of those conversations oh
48:24look at the time i gotta go
48:28well anyone have anything final to add
48:31before we before we sign off
48:36well one thing then paying digital lego
48:44on what we're just talking about the
48:46ability to automate the capture process
48:49to automate the registration to
48:52push that data into the cloud and do
48:54that as kind of a daily
48:56check of the site that where you can
48:58have robotics that can scan your project
49:01can register can push it to the cloud
49:03and perform those that analysis and see
49:07those pieces are just about there
49:09they're they're very close um
49:11and it's just to that
49:13to any new technology the key is how
49:16efficiently we can get that data
49:18available and then it allows us to use
49:20these different form factors
49:22ar when it first came out
49:24was extremely limited just because you
49:27small computing power within that
49:31within the hololens device or whichever
49:33device you're using we can see now with
49:35efficiencies in the modeling processes
49:37we can get those models
49:40much smaller in size the compression
49:42allows it to be seen on devices such as
49:45as the device does improve as the
49:47accuracy improves with these different
49:51uh it's it's definitely gonna be
49:53deployable the big question is how we
49:56make it more efficient so like the the
49:58scan to bend process
49:59if it's taking a month then it's it's
50:02not going to be viable in most use cases
50:05whereas if we find ways using artificial
50:09to make that process more proficient
50:11repeatable and precise
50:14then it allows us to with great
50:16confidence to deploy this to our
50:18different technologies and it becomes
50:20part of our workflow because it's
50:22something that's simple and adds value
50:23ultimately somebody has to pay for that
50:25month somebody has to it's got to come
50:27out of somebody's pocket hector
50:31yeah i was going to say you know we're
50:32paying attention to 5g and all those 5g
50:35towers are going up there and all the
50:36new technology is being equipped in self
50:39driverless vehicles that as far as i
50:42understand are basically scanning as
50:44they're driving along and getting all
50:46kind of data about the built environment
50:48around the vehicle and that's happening
50:51hour after hour day after day week after
50:53week year of the year
50:55i can just see um that's a massive
50:56amount of information is being collected
50:59uh of the bill of the built-up
51:01environment around us
51:03somehow we're going to have to leverage
51:05this kind of data and
51:07i'd like to tap into that at some point
51:09and so i see there's a there's a big
51:11wave of big data coming from autonomous
51:14vehicles whether those are drones flying
51:17you know uh uber helicopters are going
51:20around picking people up in the future
51:22they're collecting data of the built
51:26that's been repeated on a constant basis
51:28i think if we can tap into that and
51:31optimize that it could be a very
51:34use of lidar data that could be
51:38excellent so i think that's something
51:41well gentlemen i would like to thank all
51:44of you for your time and giving us
51:45feedback on this subject i'm sure the
51:47oda development team and the the members
51:50will greatly appreciate your input as
51:52they look to tackle this issue
51:55hope you all have a good rest of the the
51:57day and uh look to look forward to
52:00talking to you all again hopefully
52:02in person one of these days
52:09thank you everybody thank you
