Artwork for podcast Confluence
AR for Construction with Sitelink
Episode 210th July 2024 • Confluence • Evan Troxel & Randall Stevens
00:00:00 00:56:56

Share Episode

Shownotes

Jesse Vander Does and Adrian Hatch of Sitelink join the Confluence podcast to talk about their work in the construction tech industry.

Sitelink is developing a location-aware augmented reality (AR) BIM viewer for iOS devices. This tool helps general contractors and field workers access and interact with information rich building information models in real-world construction environments.

Today we go behind the scenes of the current development of their app and cover the history, challenges, and future outlook of AR technology in construction, the importance of real-time data, and technology adoption on job sites.

We also talk about how Sitelink is making BIM more intuitive and accessible, as well as their recent developments in model viewing and AR capabilities ultimately bridging the gap between design and construction to minimize rework and improve communication on job sites.

Episode Links:

Watch this episode on YouTube or Spotify.

-----

The Confluence podcast is a collaboration between TRXL and AVAIL, and is produced by TRXL Media.

Transcripts

Speaker:

Randall Stevens | AVAIL: welcome

to another Confluence podcast.

2

:

My name is Randall Stevens.

3

:

I've got Evan as usual with me.

4

:

And we have two guests today, uh,

from a company called sitelink.

5

:

ai.

6

:

So welcome Adrian and Jesse.

7

:

Jesse: Thank you.

8

:

Good to be with you today.

9

:

Randall Stevens | AVAIL: I don't know.

10

:

I'm going to.

11

:

Claim a little bit of ignorance because

your company is relatively new to me.

12

:

Um, you know, we, uh, both Evan and I

spend most of our time in the architecture

13

:

and engineering side of the world.

14

:

And you are kind of a little bit

downstream into the, a lot of

15

:

your customers in the construction

and VDC side of the world.

16

:

So we'll, we'll kind of delve into,

into that side of it, but I'll just

17

:

start this by saying, you know, one

of the nice things about doing this

18

:

podcast is we get introduced to a

lot of, uh, People like yourself and

19

:

interesting projects that are going on.

20

:

And this is a perfect example of that.

21

:

Sam from Yiga Tech introduced Adrian

and I probably what month, month,

22

:

six weeks ago, and we had some nice

conversations just talking about

23

:

technology and where all this fits.

24

:

So, uh, we're glad that you all

were able to make it on the podcast.

25

:

Jesse: Yeah.

26

:

Uh, we're excited to be

chatting with you guys.

27

:

Adrian Hatch: Yeah, thanks

for having us, Randall.

28

:

Randall Stevens | AVAIL: Yeah, I know,

Adrian, you're out, you're out on the West

29

:

Coast, if I remember correctly, and, uh,

and Jesse was talking, we were talking

30

:

just before the podcast here, up in,

uh, in Illinois, just south of Chicago,

31

:

so, um, Like a lot of companies now,

everybody's spread, spread all around the

32

:

country and collaborating on these things.

33

:

But, and Jesse, I think you said you've,

you've recently kind of switched hats,

34

:

but maybe you all can just give a kind

of general intro about yourselves,

35

:

background, and then, uh, let's,

let's dive into what SiteLink is up

36

:

Jesse: say I'm, I am well south of

Chicago, two and a half hours south.

37

:

Uh, unfortunately we can't really,

I can't just dip in, but, uh, Uh,

38

:

yeah, so Champaign, uh, Urbana,

Illinois, um, that's where I'm located.

39

:

And like you said, our team

is, uh, all around the country.

40

:

Um, and, uh, yeah, so

I'm CEO of Sightlink.

41

:

Uh, we do a location aware BIM viewer.

42

:

Um, it runs on mobile

devices, iPhone, iPad.

43

:

Um, and you can, uh, get your 3D plans

into the field and it lets you view,

44

:

uh, view your plans really intuitively,

uh, and really rapidly get answers to.

45

:

questions you have about, um,

about what you're building.

46

:

Um, our team is, how, how

big are we now, Adrian?

47

:

We're,

48

:

Adrian Hatch: No, we're on

49

:

Jesse: think a total of seven Yeah.

50

:

Randall Stevens | AVAIL: Good size.

51

:

Stay right

52

:

there.

53

:

That's the perfect size.

54

:

Jesse: pretty, we're

pretty engineering heavy.

55

:

Um, but we're growing in other areas also.

56

:

Um, so That's a little bit

about, about our company.

57

:

Evan Troxel: Who is Sitelink for?

58

:

Before we go, I just want to kind of

set this up because you're, you're

59

:

obviously developing software, but

who's your audience for the software?

60

:

It sounds like it's contractors.

61

:

Jesse: Yeah.

62

:

So I'd say the main customer

are general contractors and, uh,

63

:

it, it, uh, gets used, you know,

downstream through subcontractors.

64

:

Um, and, uh, you know, we're, We are,

you know, aimed at users in the field,

65

:

although there's, you know, a relationship

here between those in the office that

66

:

are producing the plans, VDC engineers,

and, and the, the teams in the field that

67

:

are trying to increasingly get access to

them so that they can have more agency.

68

:

Um, and, and there's an additional story

beyond the BIM viewing where we can

69

:

also take that, the context of where

you are in the BIM and let you capture,

70

:

uh, issues, photos, uh, and scans all,

uh, all, you know, properly located

71

:

within the, the, uh, um, coordinated

system so that people can understand

72

:

if there are issues in the field.

73

:

Um, but yeah, the, the main

user are, are people in the

74

:

field trying to build things.

75

:

Thanks.

76

:

Evan Troxel: Gotcha.

77

:

Thanks for, thanks for that.

78

:

I appreciate that.

79

:

Jesse: Yep.

80

:

Evan Troxel: Continue.

81

:

Go for it, Adrian.

82

:

Give us your background.

83

:

Adrian Hatch: Yeah, sure.

84

:

So I'm a chief operating

officer at SiteLink.

85

:

I've been working with augmented

reality specifically for

86

:

AEC Tech since about 2017.

87

:

And I've kind of seen that story progress

from big, high initial investment,

88

:

hardware dependent stuff on hardware

like the Microsoft HoloLens platform.

89

:

Uh, to now when we can kind of expect

pretty much everybody out on the

90

:

job site, even if they don't have,

you know, a pro Apple device with

91

:

a LIDAR scanner built in, is at

least able to, uh, utilize basic AR.

92

:

Um, and yeah, here at SiteLink have

worn a bunch of hats over the company's

93

:

history, uh, and, and these days and in

the COO seat, which encompasses, you know,

94

:

a lot of wheel on the bus type stuff.

95

:

And then also, uh, on the small

team means I still get to see a lot

96

:

of customers and users, uh, spend

some time on job sites and support.

97

:

the pipeline delivery side of the product.

98

:

So as part of that role, I'm kind of

the build manager and dev ops side

99

:

and help make sure what the engineers

are building every week makes it

100

:

out to the site as fast as possible.

101

:

Randall Stevens | AVAIL: So before we dig

into some of the features, I know we're

102

:

going to talk, kind of focus on some of

the new things that you're working on

103

:

and how that came about, but maybe you

can give a little bit of background just

104

:

about, you know, how the, how the, you

know, usually companies hopefully are

105

:

formed because there's a, you know, a

problem that now there's a solution to

106

:

solve, but maybe you can give us a little

background about how that all came about.

107

:

Adrian Hatch: Yeah, definitely.

108

:

I think I'd also be interested to hear

Jesse weigh in on this too, because he

109

:

was previously at another firm working on

some of the same problems, and we kind of.

110

:

Each of our firms approached this from

a different angle and then wound up

111

:

coming together, uh, some of those

team members, some of our team members,

112

:

uh, from the previous product, uh,

to build what is SiteLink today.

113

:

So, you know circa 2017, 2018, I got

connected with some folks on the senior

114

:

management side of large cap, uh,

commercial construction projects out

115

:

here on the West Coast and kept hearing

this drumbeat of, uh, communication

116

:

challenges affecting job site delivery

that if you drive down and really drill

117

:

down into the details, and you look at.

118

:

What's costing projects

and a lot of it's rework.

119

:

What's causing those issues.

120

:

And a lot of rework is caused

by communication failures.

121

:

You can see there's a clear story that

having the people back in the back office

122

:

and the managers, even the folks who

are onsite many days of the week, but

123

:

are on different sites, keeping them

in the loop with the folks, uh, who

124

:

are out in the field doing the building

and then vice versa, making sure those

125

:

folks out in the field have the latest

and greatest on, on the design side.

126

:

That's a big challenge for this industry.

127

:

It certainly was then, and then even

now today as more people are bringing

128

:

digital communication tools out, it's

really far from solved right now.

129

:

And we looked at how can we use augmented

reality to enable that, to let somebody

130

:

out on site capture details about an

issue they're seeing with the context

131

:

of where they are, have them get better

access to those 3D BIM plans that doesn't

132

:

rely on making it back to the trailer,

Uh, and, and cracking into the BIM rig

133

:

that we have in the trailer that lets me

actually view the plans from my building.

134

:

Um, and I know simultaneous to that

at the project Jesse was working on,

135

:

they were working very hard on that BIM

overlay location aware access to plans.

136

:

Well, we were working pretty hard

on let me make markup out in the

137

:

field that lets me see what's

going on in context of the site.

138

:

And so link kind of brings those.

139

:

to solutions to a common problem together.

140

:

Now we've got a tool that gives you

that real time, quick access to that

141

:

BIM model, lets you see it on your

screen, lets you see it overlaid

142

:

onto the real job site environment.

143

:

But then when you see that issue, or

you identify that item that needs to

144

:

go back to the back office, maybe it's

a great thing, maybe you just did your

145

:

inspection and everything's all good.

146

:

You can capture that documentation with

the context of the real site, capture

147

:

it as a scan, capture it as a photo,

or But make it a part of that project's

148

:

digital project management story, put

it into that BIM model, so whoever needs

149

:

to see it back at the office has access

to it, uh, with the context they need.

150

:

Heh heh heh

151

:

Randall Stevens | AVAIL: know we talked,

when we talked, uh, the other week,

152

:

Adrian, we were, you know, I was asking

about how much just the evolution of

153

:

the, in this case, the iPhone hardware,

kind of the consumer level, everybody's

154

:

got one in their pocket, but, you

know, when, one, is it accurate enough

155

:

without localizing for y'all to do

what you need to do and, or how do you

156

:

augment that when you get on site in

order to get the level of accuracy?

157

:

And, and then when did that, yeah.

158

:

When did that happen?

159

:

Right?

160

:

you know, when did all the

technology conspire to come

161

:

together to be able to actually do

these kinds of things in reality?

162

:

No pun

163

:

Jesse: we get that question

a lot, uh, accuracy is, is a

164

:

very big question for people.

165

:

I, I'll, if I could come back to it

though, cause there are other aspects

166

:

of the, like commodity hardware that

I think are definitely were part of

167

:

the observation we were having, which

is that, uh, the, the iPhones in

168

:

particular Their rendering performance,

uh, is really pretty incredible.

169

:

Uh, it's not to say you can just take

a, you know, model designed to run on

170

:

a gaming rig and just drop it onto an

iPhone, but the, the ceiling for what you

171

:

can show a user has lifted considerably.

172

:

Um, and the, you know, uh, like Adrian

was saying earlier, the availability of

173

:

those devices is, is massive, Right, So.

174

:

That's one observation.

175

:

you add in the proliferation of LIDAR.

176

:

Uh, and that's another one of

these things that previously, like

177

:

you just wouldn't have thought of

having on a consumer grade device.

178

:

Um, and while certainly not say

industrial grade, it still has a

179

:

lot of information to share, right.

180

:

And then finally, um, uh, ARKit

augmented reality, uh, it gives you the

181

:

ability to track where users devices.

182

:

Three dimensionally, um, with, you

know, again, just it's available.

183

:

It's just on your phone.

184

:

So all of those things though are

consumer grade technology, right?

185

:

They're not meant, you know, they're not

designed with the industrial use cases

186

:

in mind, but the question for us kind of

fundamentally is, to what extent does the

187

:

kind of broad access to that make, uh, you

know, really change the game in terms of

188

:

how we can provide information to users?

189

:

And the accuracy question, We kind of

have to flip it on its head because

190

:

the short version of it is that like

AR kit, for example, on its own is

191

:

not accurate enough, for example, to

like point your phone at something and

192

:

be like, I'm going to drill the hole

193

:

Randall Stevens | AVAIL: Right, right.

194

:

You're not manufacturing

with that info, right?

195

:

Jesse: no, no.

196

:

And, and I, Personally, I

don't really see a future where

197

:

that is what you do with it.

198

:

Um, but what you can do with it

is change how people interact

199

:

with their existing plans, Right.

200

:

And the kind of exercise I go through is

like, how do you currently use your plans?

201

:

Right.

202

:

Do you print it out world size and,

and try to drill holes with it?

203

:

Or are you looking at it and

trying to understand it relative to

204

:

other things in your environment?

205

:

And that's where, uh,

SiteLink really excels.

206

:

Because what we can do is, uh, so

here's a very concrete example, right?

207

:

We have a user who, uh, put drywall

up and needs to see what is behind

208

:

the drywall in a particular stud bay.

209

:

And I kind of challenged them, I asked,

well why wouldn't you just use like

210

:

a regular 3D, you know, BIM viewer?

211

:

You can get that answer by looking at

a 3D model on an iPad in a BIM viewer.

212

:

And their answer is simply like with,

with SiteLink, you just point your

213

:

phone up and you look at it and you

can see what is behind that stud bay.

214

:

Right.

215

:

Uh, and so that isn't about

getting one inch accuracy.

216

:

That's about making it much

easier for you to understand the

217

:

information that is already available

218

:

Randall Stevens | AVAIL: Yeah, you're

219

:

Jesse: And so that's,

220

:

Randall Stevens | AVAIL: contextualizing

it to the environment that you're

221

:

standing there in, you know.

222

:

Right.

223

:

Evan Troxel: quick stories I want

to share that super fast here.

224

:

The first one was, uh, having a

conversation with Nick Cameron at

225

:

Perkins and Will about accuracy

with, with scanning different, you

226

:

know, Matterport as an example with,

where it's using infrared scanning

227

:

instead of LiDAR scanning, right?

228

:

And the question was like, I think in

their terminology, at least back in the

229

:

day, it was like 99 percent accurate.

230

:

And, and somebody will say, well, in

eight feet, that's an inch, right?

231

:

That adds up.

232

:

It's, it's, it's a, it adds up to a lot.

233

:

And, and, and Nick.

234

:

Push back.

235

:

He challenged, he goes, it's

more accurate than an intern.

236

:

Jesse: Mm

237

:

Randall Stevens | AVAIL: Nick

238

:

Evan Troxel: and how many times will you

have to send the intern to the job site

239

:

to remeasure or to measure correctly for

the third time and take the right photos?

240

:

Right.

241

:

And so it was like, it's accurate enough.

242

:

And, and I, and then I think about.

243

:

The contracting world, right?

244

:

Like we, as an architect,

we're drawing really precise

245

:

stuff.

246

:

It will not get built that precisely

for the most part, in most cases in like

247

:

what, 98 percent of the cases, it's not

even going to, you know, the tolerances

248

:

are much different in the real world

than they are in the digital world.

249

:

And, and I know that depends on the

project and the technology and all

250

:

those things, but the funny story

is I was building a pool at my house

251

:

and I went out to stake the design on

the ground with the pool contractor.

252

:

And I'm like, okay, this leg of the pool

is like 11 foot four and three quarters.

253

:

And he's like, no whole

foot increments only.

254

:

That's what he told me.

255

:

And I was like, what

are you talking about?

256

:

It's like, that's all my crew does,

257

:

you know?

258

:

And I was like, that was wild.

259

:

That was an eye opening

experience for me as an architect.

260

:

Jesse: they have specialized

tape measures for

261

:

Evan Troxel: Yeah, that's what I was

262

:

wondering.

263

:

I'm

264

:

Jesse: yeah,

265

:

Evan Troxel: are all the, what

are all the other numbers for?

266

:

Jesse: blank tape measures,

267

:

Evan Troxel: Right.

268

:

Randall Stevens | AVAIL: But I

269

:

think, I think, uh, you know, another

way to think about this is that any

270

:

and all of this technology still

does not replace good judgment.

271

:

People on, in the situation, aware of

what that is, use the information, but

272

:

there's always human judgment about

273

:

Evan Troxel: verify.

274

:

Yeah.

275

:

Jesse: which I think is part of, I think

that's part of a design challenge for

276

:

us, you know, cause I think that, uh, you

know, especially, you know, we're not the

277

:

first company to do a BIM overlay, right?

278

:

Like a camera feed with your

plan superimposed on top of it.

279

:

And when you, as soon as you go there, I

do think that there is a, there are trust

280

:

issues and there are certain expectations

that come, you know, that people.

281

:

suddenly have, Right.

282

:

Like they want it to be

eighth inch accurate.

283

:

And I, think it's an ongoing area

that we're exploring because obviously

284

:

the more accurate, the better.

285

:

Uh, it helps you understand it more

deeply, uh, and trust wise, it just

286

:

makes you feel comfortable that the

system's performing as it should be.

287

:

Uh, but all of those things

are interesting design

288

:

areas for us to explore.

289

:

Like, you know, if it isn't

working well, can we tell you

290

:

so that you know what to expect?

291

:

Uh, and are there ways for us to

convey like, yeah, that this is not

292

:

intended to be install accurate.

293

:

This is intended to be, uh,

you know, for you accessing and

294

:

understanding the documents.

295

:

Um, and those are, I think

somewhat open questions for us.

296

:

Adrian Hatch: I, I think

297

:

Evan Troxel: constantly

evolving, I would assume.

298

:

Yeah.

299

:

Adrian Hatch: That human judgment side,

I think, touches on one of the core

300

:

values that's, that's come through in

this product's feature design, where

301

:

it, like we've mentioned a couple of

times, it's not just that BIM viewing

302

:

capability, but that ability to, to call

in resources from outside your site.

303

:

To capture the current state of the site

within the model, uh, because everybody

304

:

who's worked on a job has, has experienced

the stories where everybody thought they

305

:

were making the right judgment call with

the available information on their desks

306

:

or in their trailer, uh, but everybody

was missing some key piece of context.

307

:

And so encouraging folks, not to just let

the model, not to let the plans do the

308

:

autopiloting, but to really collaborate.

309

:

with one another, with more richer,

more accurate context is, is

310

:

really a key part of the story.

311

:

Randall Stevens | AVAIL: I know, uh,

Adrian, I think you were going to share

312

:

some, um, you're working on a new, uh, Uh,

Model Viewer feature that's part of this.

313

:

Is that something you can end

up showing us some of what,

314

:

what we're describing verbally?

315

:

It's always easier to,

for everybody to see this.

316

:

Adrian Hatch: Yeah, Jesse, how about

I, I run us through a couple slides of,

317

:

like, actual overlay site link making

that kind of sets the table on what this.

318

:

does right now, and then we can

kind of roll it over to you to show,

319

:

uh, what we've got coming up next.

320

:

So,

321

:

Jesse: like that.

322

:

Adrian Hatch: And, uh, as I promised

everybody on sales calls, I keep

323

:

my slide decks as fast as possible.

324

:

So you're not even going to get slide one.

325

:

We're like on slide 10 here already,

and we're going to be quick.

326

:

So we've already belabored

All of these points.

327

:

This is what it does.

328

:

You know, it's a collaboration tool.

329

:

You can see that happening

right here already.

330

:

We've pointed out a broken fitting that

we're just sharing right within iOS.

331

:

Like one of the key values of

this platform is not to add.

332

:

More duplicate tools to your workflow.

333

:

So you're gonna see all our sharing

goes through the tools We already

334

:

use to contact one another on site.

335

:

iMessage is like the number one on site

collaboration tool already Right here.

336

:

You see we're looking at a BIM model

out on a site This does integrate

337

:

to Autodesk Construction Cloud so

you can bring those models right.

338

:

in from your ACC project This is

that BIM overlay that we're familiar,

339

:

you know, I think this is a familiar

concept in this business at this point.

340

:

Choose, uh, from your available views,

see those views individually or in

341

:

combination, punch through the walls with

X Ray, or see it, you know, the way it

342

:

would be if it was installed overhead.

343

:

This is a little more of that.

344

:

You can see there's that accuracy level

Jesse was talking about, where you

345

:

wouldn't necessarily install these pipes

exactly against what's there in the model.

346

:

But it's, it's close enough that you're

very genuinely seeing an accurate

347

:

representation of reality on site.

348

:

Um, and over here, we're creating

documentation out in the field.

349

:

That then we're able to send

and share with other folks on

350

:

our team, however we need to.

351

:

Uh, it can capture 3D scans while

you're doing that, while you're creating

352

:

that documentation out on the site.

353

:

Uh, Wayfinder way through the job, maybe

to even something somebody placed in

354

:

that viewer off the site themselves.

355

:

Uh, so that's, that's

where it's at right now.

356

:

You know, it allows you to capture this

documentation, share it externally,

357

:

view your plans on the site and overlay.

358

:

And what I'll let Jesse take it over

with here is How that story is now,

359

:

the requests went beyond, Yeah, let

me see what's on my site right now,

360

:

let me see that model overlaid.

361

:

The job to, let me get a BIM viewer

that's as convenient as that.

362

:

Like it's, it's this easy to

just see what's in that wall.

363

:

Let me just see the model for my

building, uh, with the same ease

364

:

of use and the same, uh, you know,

efficient access to information.

365

:

And, and that's where we started

working on this standalone model viewer.

366

:

And when we were preparing to get on

this call today, and I thought about

367

:

telling that story of how a concept

for a solution around a problem becomes

368

:

a solution suitable to the industry.

369

:

How this feature kind of came together

has changed a little bit and, and what

370

:

we're releasing now is going to be a

little different from what we'd been

371

:

originally piloting with our users.

372

:

Jesse: That's a very

cool looking app, Adrian.

373

:

Adrian Hatch: Yeah, try it.

374

:

Jesse: Um, Yeah, so I,

that should we share the,

375

:

Adrian Hatch: Yeah, go right

376

:

Jesse: variation thereof?

377

:

Adrian Hatch: yeah,

378

:

let's, let's show that

379

:

Jesse: unless you guys have any

questions about that feature set, I mean,

380

:

Adrian kind of, that, that was a very

quick overview of, uh, Uh, what's in

381

:

Evan Troxel: have a quick,

just a quick question.

382

:

How are you doing the location?

383

:

Is it, because I've seen

384

:

different ways of doing that, you

know, back in the old days with

385

:

an iPad and you could scan a QR

code or something, and it would

386

:

use

387

:

Jesse: it is.

388

:

You got it,

389

:

Evan Troxel: or, or if you're using GPS,

I can't imagine GPS would be accurate

390

:

enough, especially with all the, the

buildings in the way potentially.

391

:

So yeah, we'd love to hear about that.

392

:

Got

393

:

Jesse: So we use QR codes.

394

:

Um, so essentially what, what we have is,

uh, um, a, uh, uh, a web portal where you

395

:

can, uh, tell us where a QR code will be

on a job site, um, and, and, and This is

396

:

an area that we're improving on currently.

397

:

And I'm, I'm excited for where

we're going here because, uh, well,

398

:

there are two parts to this, right?

399

:

Basically set up of these systems is

one of the big areas where users get

400

:

tripped up and then don't make it.

401

:

to the big value.

402

:

Right.

403

:

And so when we started

doing this, we were.

404

:

very aware of wanting the setup

to be as easy as possible.

405

:

And like you said, there are

a lot of different options.

406

:

Uh, QR codes can be challenging

because you want to have the

407

:

right QR code in the right place.

408

:

And then you're, you know, some,

uh, systems have you like model in

409

:

a QR code and you have to have the.

410

:

ID of the QR code in the model and then

make sure that's the right one in the

411

:

right place on the physical job site.

412

:

So the way ours works is you go, you put

up QR codes, you actually walk around with

413

:

the phone and scan them in and we get a

3D location of each of them and then you

414

:

can drag and align them into your model

and then you can fine tune from there.

415

:

It's pretty good, uh, it's, it is a,

you know, one of the faster setups,

416

:

uh, maybe the fastest setup in the

industry, but we're not there yet.

417

:

Um, so,

418

:

of the things we're exploring are sort

of like a hybrid approach, um, where

419

:

we could, uh, for example, let you

indicate that this object is a QR code.

420

:

Sorry.

421

:

Like it is a QR code, but it doesn't

need to be the specific QR code.

422

:

And then we can, uh, potentially

automatically align based

423

:

on some hints you give us.

424

:

Uh, the goal there is to, try to

loosen the, like the connection between

425

:

the modeling that you have to do and

that onsite setup, um, and move as

426

:

much of the setup as possible onto

the mobile device so that you don't

427

:

have as much dependencies on the

office to get, get up and running.

428

:

Um,

429

:

Randall Stevens | AVAIL: far off are

we, Jesse, from being able to just.

430

:

recognize the, what's on the

site and begin to align to it.

431

:

Right.

432

:

Um, like when, in those videos, when

you're showing that there's already

433

:

framing that's going up, you know, if

you knew that, you know, can you, can

434

:

you begin to, is there a future where

you can just begin to automatically

435

:

recognize and do automatic alignment?

436

:

Jesse: I don't want to say,

437

:

think we're a ways off from it.

438

:

Um, I think that, uh, it's a job

sites are a particularly challenging

439

:

environment for computer vision.

440

:

Um, they change at a rate that

is really hard for, uh, for the,

441

:

class of algorithms that are, are

used for doing this sort of thing.

442

:

So for example, just built into ARKit,

um, and into the HoloLens and into

443

:

ARCore, right, all of these AR systems.

444

:

Um, and actually for that matter, any of

your photo reconstruction systems have

445

:

a way to say, is this a place I've been

to before and then relocalize to that.

446

:

But it presumes that there's enough

commonality between that and the,

447

:

And the last time You visited it.

448

:

Right.

449

:

And by definition, a job site

is changing really rapidly.

450

:

Randall Stevens | AVAIL: so.

451

:

Yeah.

452

:

Jesse: Yeah.

453

:

Yeah.

454

:

Um, but like you said, I think the, you

know, there's, it's tantalizing because

455

:

you also have these plans that are saying,

this is what is supposed to be there,

456

:

and then you observe what is there and

theoretically, could we line up to that?

457

:

I, I think, I think, yeah, I think

there's a chance that we can get there.

458

:

Um, There are a couple of parts to that.

459

:

One is that like the as built versus the

plans, that's sort of the whole point

460

:

is that it, it isn't always the same.

461

:

Right.

462

:

So lining up to that may or may

not be what you actually want.

463

:

Right.

464

:

Um, and the other thing I'll say

is I you know, again, I think more

465

:

about how to sort of augment a

user's existing kind of knowledge.

466

:

So one thing we've explored, we were

actually prototyping this last week.

467

:

Um, if you tell us, you know,

uh, roughly where you are.

468

:

And we should be able to refine

the viewpoint to kind of snap you

469

:

into an even better, uh, alignment.

470

:

So that's where my head's

at with, with that.

471

:

But I, I wouldn't say that I'm super

bullish on a sort of magical, like point

472

:

your camera and we will find what floor,

what room and what stage you're on.

473

:

Um, I think it's, that feels

like a very challenging problem.

474

:

yeah, those are good questions.

475

:

All right.

476

:

Well, with that, how about the,

how about this, uh, this, uh,

477

:

feature that we've been working on?

478

:

Um, so yeah, just the, some of the

features you guys saw the way I've been

479

:

kind of describing it by way of metaphor

is that we want to be sort of the Google

480

:

map of BIM, uh, and when we look at

BIM overlay, uh, what others have done,

481

:

there tends to be this sort of, you are

either BIM overlay or you're, you're

482

:

Uh, kind of traditional viewer, and

we're looking to explore an interaction.

483

:

That's a lot more fluid and, and blurs the

line into what we call location aware BIM.

484

:

And so, uh, the idea here is that

you, um, should be able to, uh, simply

485

:

pinch and pull out, look at your model

and then hit locate and come back.

486

:

Uh, and so, Um, so here I am.

487

:

This is not a great example cause

it's just in my house, but you'll

488

:

see, this is AR mode, traditional

AR mode, very roughly aligned.

489

:

Uh, but then I can pinch zoom and pull

out and, uh, we can look around just as,

490

:

uh, you would in a traditional BIM viewer.

491

:

I want to point out that like, you

know, when we talk about the value of

492

:

location, so like with BIM overlay.

493

:

You can say, look at that stud wall

and understand what's behind it.

494

:

A view like this, you actually

can accomplish a very similar

495

:

thing where you could.

496

:

Take your phone, which is that

blue dot and walk up to the, uh,

497

:

to the wall and see which section

of the wall in your model you're

498

:

actually pointing your phone at.

499

:

Right.

500

:

Um, and so the awareness that like

spatial awareness is really powerful,

501

:

but now you can just navigate the

model like you would traditionally.

502

:

And then just like Google maps,

you could hit your locate button

503

:

and it will bring you right back

in to your BIM overlay mode.

504

:

Yeah, that's the, uh, that's

the feature we wanted to share.

505

:

Evan Troxel: Nice.

506

:

So, so you're basically, the analogy

is like street view to map view, right?

507

:

And you're, you're going in, you're,

you're coming back out and that

508

:

gives you kind of that, the, the

different, the two different levels

509

:

of the map that you can experience.

510

:

Jesse: Yeah.

511

:

And I, you know, it's Street

View is a good analogy.

512

:

I actually was even just

thinking when you're running

513

:

like GPS for your car, Right.

514

:

uh, you have those two modes.

515

:

One is you're locked onto that blue dot.

516

:

Uh, it's oriented by the compass, right?

517

:

And, and you don't even think about it.

518

:

You just, if you want to go look at

something else, you just drag, right?

519

:

Uh, those modes are

kind of really blended.

520

:

Uh, and that's what we.

521

:

That's what, uh, we're going for

522

:

Evan Troxel: Nice.

523

:

Yeah, that's great.

524

:

Okay, I get it.

525

:

Jesse: Yeah.

526

:

Adrian Hatch: I think that's what I was

527

:

Jesse: I think, yeah, go ahead.

528

:

Adrian Hatch: the design process

is that That's just, you know, what

529

:

we've gotten to now, and we're now

getting ready to release when we first

530

:

started thinking about that feature.

531

:

We kept, I, you know, it was as

simple as, oh, and then you'll, you'll

532

:

switch to model view mode, and you

won't be, like, localized anymore.

533

:

You'll be looking around the model, and

then you'll hit a button, and you'll

534

:

come back to AR mode, and I think, uh,

it was from working with users and from,

535

:

and from knowing the, that, that compass

this company has of location aware BIM

536

:

was, wait, this really is one mode.

537

:

You're still out on the site.

538

:

You're still getting the job done.

539

:

Why toss out that location context

when instead we could make that

540

:

useful and make a BIM viewer that has

some capabilities, uh, nobody else

541

:

has at this time while we're at it.

542

:

Randall Stevens | AVAIL: So when you're,

when you're in the, uh, overlay mode

543

:

on site, like in the examples you were

showing the video where you're seeing,

544

:

you know, piping or conduit or, or,

um, as you would pinch, you would still

545

:

just be seeing that layer or, and then

you'd have the option to turn on more

546

:

of the model if you wanted to see it.

547

:

Is that the idea?

548

:

Jesse: Yeah, I think as much as

possible, those two views will

549

:

operate as identically as possible.

550

:

That's another part of

it is just to make Sure.

551

:

that there's just one sort

of style of interaction.

552

:

So you want to see multiple, you know,

you want to see a different layer.

553

:

It's the same process you would elsewhere.

554

:

Um, yeah.

555

:

Evan Troxel: I feel like filling

in the blanks between those two

556

:

different perspectives is huge from

a user experience because it gives

557

:

me the contextual awareness that I

need to understand what's happening.

558

:

Whereas, how much software do

we use nowadays where switching

559

:

between interfaces, we get no

clue what's happening, right?

560

:

I grew up in the days of Photoshop in the

90s where it would prepare to save, right?

561

:

And we had the the little

progress bar would march and

562

:

then, and then it would save.

563

:

And then we could watch it again.

564

:

Right.

565

:

And basically you could get up and go get

a cup of coffee while it did all that.

566

:

But at least I knew what it was doing.

567

:

Right.

568

:

And, and to me, there's a lot of

software nowadays where we just

569

:

have no idea what it's doing.

570

:

And it's not giving us any kind

of breakdown of expectations.

571

:

Right.

572

:

And so by, I think the user experience by

just adding that transition between those

573

:

two different perspectives is, is huge.

574

:

And I think people will just take it

for granted and they won't know why.

575

:

It just makes sense.

576

:

And they won't know why it's

kind of, you know, this, this is.

577

:

Kind of fun, but if it was

missing, it would be, it would

578

:

be like where I just got lost.

579

:

Where am I right?

580

:

Like people can't read plans as

it is when I'm layman, right?

581

:

If you're, if I'm talking to a

client and I'm showing them floor

582

:

plans, the orientation is very

difficult because it's not reality.

583

:

It's not 3d.

584

:

It's not the right scale.

585

:

It's none of those things.

586

:

And.

587

:

When I put them in a VR environment,

right, fully immersive, it's like,

588

:

okay, now I understand this, if I

can get them to put on a headset,

589

:

right, but there's still kind of

these, these barriers between it.

590

:

So if I've got this device, I'm super

familiar with, I've, I've got, I'm used to

591

:

looking at images, if I can hold it up and

I see it as an overlay using this magic

592

:

window of sorts to just see what's going

on behind, you know, right there, right

593

:

on the other side of that sheet of glass.

594

:

But then I can transition to a

dollhouse view and, and see that happen.

595

:

I, I I think that that is

a, is a really big deal.

596

:

Jesse: think, yeah, thanks for saying so.

597

:

I think, you know, the, um, you've hit on

something there that like, when we were

598

:

first playing with this, I think there's,

we have to do a little bit of a gut

599

:

check where it's like, Is this just cool?

600

:

Like, cause there's, it's a fun demo.

601

:

Evan Troxel: Is it just eye candy?

602

:

Right.

603

:

Jesse: exactly.

604

:

Like, like, Whoa, like that's yeah.

605

:

Neat.

606

:

Uh, but a couple of things on the

transition that you mentioned, Right.

607

:

So in that video, I actually just play

it backwards to transition back in.

608

:

Um, it, and the first version of it, you

hit the button and it snaps you back.

609

:

To the original view and those,

like exactly as you said, right,

610

:

that snapback is jarring, right?

611

:

Like the transition back into the

camera view is not just cool, it's

612

:

what your brain needs able to transit

to, to come along for the ride, right?

613

:

Um, and so those are things

that we, you know, we.

614

:

Yeah, I, I think.

615

:

we're exploring with it and in

some, sometimes it's sort of

616

:

obvious, like, yeah, like, why

wouldn't you do it that way?

617

:

But sometimes we have to build it and

then play with it and go like, oh yeah,

618

:

that's not as good as it could be.

619

:

Um,

620

:

Randall Stevens | AVAIL: I was

talking to somebody recently.

621

:

I can't remember who the conversation

was, was with, but it was kind of,

622

:

the conversation was something along

the lines of AR was kind of hot five

623

:

or six years ago, which we all know

it kind of, you know, hype, the hype

624

:

cycle, you know, went blah, blah, blah.

625

:

What's, and I, I recently been renovating

a condo myself and I was using it, right.

626

:

To start to visualize and.

627

:

Not only for myself, but my wife

understands some of the things that I

628

:

was thinking and designing in there.

629

:

I think for people who haven't looked

at it in the last four or five, six

630

:

years, when the hype cycle, and then

they said, Oh, that's kind of cool

631

:

and passe, it has gotten really good.

632

:

I think the AR kit and the ease with

which you can now, the scanners, right?

633

:

If you haven't used any of the scanners

to start to, you know, it's like four.

634

:

A lot of very useful use cases.

635

:

It's good enough, like, to, to

use for those kinds of things.

636

:

So I, I can, that's why I think it's

interesting that on the professional

637

:

side, the last five years of the hardware

advances and, and what's been happening

638

:

with the AR kit, it's like gotten to

that point where it's like, look, this

639

:

doesn't have to be, you know, sub,

you know, 10th of an inch accurate.

640

:

It's like good enough to

start getting the work done.

641

:

And, um, so I think, um, I just

encourage anybody that hasn't taken

642

:

a look recently, just get your phone

out, go download one of these apps.

643

:

There's plenty of them, kind of prosumer,

uh, that you can start to do that.

644

:

Uh, our sister company, Archvision

has an app called Fovea.

645

:

I'll give a shout out for their

app, but you can install it and

646

:

bring your, that's what I was doing,

bringing the Revit model in that I

647

:

was laying some stuff out in Revit.

648

:

And it's like, boom, it's

like, it just magically worked.

649

:

And,

650

:

um, anyway, so just for somebody

that doesn't have my head in it

651

:

every day, uh, like some people

do, it was like, wow, this has

652

:

really gotten approachable, useful.

653

:

Um,

654

:

Jesse: totally.

655

:

I've done the scanning on my iPhone

was definitely kind of revelatory.

656

:

I had a sort of like an insulation issue.

657

:

I have a very old house, like pre 1900s

and the bathroom floor was frigid.

658

:

And I did a, uh, poly cam scan from,

The, you know, the kitchen all the

659

:

way through the first floor of the

house, cause that's the shortest path

660

:

up the stairs back to the room that's

on top of it, which is my bathroom.

661

:

And, you know, with that, I could

see that my, my bathroom actually

662

:

overhangs my back deck, right.

663

:

And that just wasn't obvious to me.

664

:

Like I, you know, I'm like, why is that?

665

:

Why is it like subzero

temperatures on this floor?

666

:

I'm like, oh, it's, it's, it's.

667

:

It's basically outside and

has no insulation on it.

668

:

And yeah.

669

:

those types of tools to be able

to do that with your, uh, with

670

:

the phone that you just use for

everything else is pretty incredible.

671

:

Randall Stevens | AVAIL: Yeah.

672

:

I haven't, uh, I haven't personally tried

the new Uh, Apple Vision Pro headsets.

673

:

But I was talking with somebody just

recently that, that was at the Apple store

674

:

and got a demo of it and they said, it's

absolutely amazing, but there is no way

675

:

that I would spend that much money, right?

676

:

Like the use, the novelties

there, but the justification,

677

:

you know, is like way up there.

678

:

I think, you know, with what we're

talking about though, it's like,

679

:

Not many of us aren't walking around

with, you know, one of these devices

680

:

for a bunch of other reasons and

now you can take advantage of it.

681

:

Evan Troxel: In the construction

business, Apple Vision Pro is nothing.

682

:

compared to a construction

683

:

project, right?

684

:

And so I, I did want to ask, like,

are you guys considering developing

685

:

because you're developing on iOS?

686

:

Are you also considering

developing for Vision OS?

687

:

Because it seems like You know, and

I'm not a developer, so it's just, it's

688

:

easy, right, to, to do that, but, but

I could imagine people will be walking

689

:

job sites with, if they've, they've been

doing it with HoloLenses for a long time,

690

:

right, and now the Vision Pro is kind of

this hybrid device where it's augmented

691

:

reality and virtual reality, and you

can kind of use the crown to dial back

692

:

and forth, but as an augmented reality

device, instead of holding up the piece

693

:

of glass, it seems like I can just look

where I want to look, and I can maybe dial

694

:

in the BIM model on top of the reality.

695

:

Randall Stevens | AVAIL: I mean,

I think the, well, Jesse, I'll

696

:

let you jump in, but I think

it's like with ArcVision's tech.

697

:

They didn't do anything to

make that anything special.

698

:

It was basically everything that

was developed to run on this

699

:

just magically works then in.

700

:

So that's the beauty of Apple and

that development platform is all

701

:

of these become prototyping devices

for now that same tech stack to

702

:

work on the Apple Vision Pro.

703

:

Jesse: We, yeah, so there is

a bit more work for us to do.

704

:

So part of what we, you know part of, What

makes SiteLink powerful is our ability

705

:

to take these really massive models and

get them running on your, your phone.

706

:

Um, and to that end, we

have a pretty custom engine.

707

:

Uh, and, um, yeah, so there's some

work we have to get it running.

708

:

Uh, and yeah, I'd say.

709

:

it's not where our focus is Right.

710

:

now, but it's definitely

an area of interest.

711

:

Um, and we're kind of letting, I

think, other people try things out.

712

:

I mean, top of my mind is like questions

of like, of safety, um, you know, and what

713

:

people, um, what people's appetites are

for pressing that boundary, like apples,

714

:

you know, uh, Official line is don't

use this anywhere that it's dangerous.

715

:

Like just don't, Right.

716

:

Um, I don't think that's going to stop

people from trying and I've talked

717

:

to people who are absolutely going

to try that, um And I actually think

718

:

it's an interesting question because

I don't know that the pass through is

719

:

say more dangerous than the HoloLens.

720

:

On the HoloLens you could put so

much holographic data in front of

721

:

a person's eyes that they couldn't

see where they were walking, Right.

722

:

So yes, it's a transparent screen, but I

don't know that it was a safe, you know,

723

:

just, you know, intuitively safe, Right.

724

:

You had to be very mindful of how much

content you were showing at a given time.

725

:

But with that said, like, you at least

couldn't just Dial it to zero, right?

726

:

There wasn't that option.

727

:

So I think there's some open

questions for putting it in the field.

728

:

Um, my initial thought on the vision

pro was about, you know, working the

729

:

trailer to collab or even further

away, remote access, where we could say

730

:

beam a, a scan from the job site back.

731

:

Uh, in some sort of like, 3d video

conferencing, you could be talking to them

732

:

in the context of the real life model.

733

:

Um, but I had the use cases kind of more

office or, or, um, you know, trailer side.

734

:

Um, but I, I mean, I think it is I want

to see what other people do with it.

735

:

Right.

736

:

I think there's in some ways,

like people do have to push that

737

:

boundary to just see what it can do.

738

:

Um,

739

:

Randall Stevens | AVAIL: So it sounded

like, is, is the limitation, Jesse, the,

740

:

uh, how, how much data you can render

and how quickly you can render it?

741

:

Is that the main

742

:

limitation?

743

:

Jesse: I, you know, I'm putting

my CTO hat on so I can get a

744

:

little bit more nitty gritty.

745

:

Right.

746

:

Uh, so our renderer is built

on something called scene kit.

747

:

And, uh, actually I was somewhat

surprised when it was released.

748

:

It, it requires, uh, that it's

rendered in reality kit, which we

749

:

specifically chose one over the other

for performance reasons at the time.

750

:

Uh, and so we would have to port, um, a

little more than as you were saying, like,

751

:

you know, I think potentially had we built

it in reality kit to begin with, then it

752

:

might literally have been like a, Yeah.

753

:

move this over, probably make a

couple adjustments and hit build.

754

:

Um, it is worth noting

though, like we've, Yeah.

755

:

we intentionally built our whole app

in, you know, it's native to iOS.

756

:

It's not a unity app.

757

:

Um, you know, the experience from

the ground up has been, we've, we've

758

:

Always had an eye for wanting this

to be a really pleasant experience.

759

:

And, you know, that goes from like opening

the app, having it launch quickly, uh,

760

:

to all of the native buttons and all the

things people expect in a modern app here.

761

:

Um, but Yeah.

762

:

so we are set up to be able

to port it, but it's just not

763

:

on our kind of top 10 list.

764

:

Evan Troxel: well, it just seems

like a legit use case where there's

765

:

a lot of apps out there who are

scratching their heads saying, like,

766

:

we have no idea what we would do.

767

:

It seems totally obvious with something

like this, and the idea of, like,

768

:

the remote viewing, I think, is legit

as well because, oftentimes, you

769

:

know, from an architectural office

standpoint, we're sending one person

770

:

to the job site meetings every week,

but there's the people who answer the

771

:

questions are back in the office, right?

772

:

They're the people with 30,

40 years experience who are

773

:

working on 20 different projects.

774

:

And if they can tap into seeing what's

happening on site, again, I think

775

:

that's, that's a really legitimate,

if they can see it in, in reality,

776

:

in my podcasting air quotes, right?

777

:

Like that is a game changer versus A

grainy dark photo that somebody snapped

778

:

and, and sent over email that they, that,

you know, along with the 50 other emails

779

:

that came in today, it just seems like

there's, there's a way to kind of bridge

780

:

the gap between the office and the site in

a, in a really amazing way there as well.

781

:

Randall Stevens | AVAIL: See,

782

:

Nick, I was going to say, Nick

is going to want to send his

783

:

intern back to the job site.

784

:

Now with one of these headsets on,

it's basically his autonomous bot.

785

:

Adrian Hatch: I was going to say,

we're just going to have to show Evan's

786

:

firm how much you can do with SiteLink

through a FaceTime call, because you

787

:

can bring a whole lot of staff members

out on site and contacts that way.

788

:

Evan Troxel: think that's

a, that's another thing that

789

:

benefits the industry, right?

790

:

And because not enough people get to

go to the site to visit the site and

791

:

see, what am I drawing versus what's

getting built and why did it matter?

792

:

And why, how could I do it better?

793

:

And the mentoring and just the kind of

the stuff that doesn't happen as much

794

:

because there's no resources to do that.

795

:

Or, you know, My, my region of a firm

is getting much larger, so sending

796

:

people out is getting less and less

of an option because we're competing

797

:

for projects farther and farther away.

798

:

So, it's, it's things like that that

are all being addressed with technology.

799

:

Jesse: Yeah.

800

:

And I think the, you know, to jump

on that theme of, uh, like commodity

801

:

hardware getting more and more

powerful, the, you know, spatial

802

:

video, I think fits into this.

803

:

Um, so the, you know, the ability for

you to record stereo video, um, and

804

:

then view it on a platform like this.

805

:

I think that's really powerful.

806

:

Uh, and again, just the

availability of it, Right.

807

:

Like if you want to describe an issue

you're having, being able to simply record

808

:

it, uh, or potentially stream it, uh, and

view it in a headset where you can, you

809

:

know, richly understand the, the 3d scene,

I think those are really good use cases.

810

:

Randall Stevens | AVAIL: Are y'all,

you know, your choice of deciding to.

811

:

Develop exclusively on the iOS platform.

812

:

That's interesting because I would

imagine if people are bringing their

813

:

own devices, you probably have at

least half or more on construction

814

:

sites that are probably Android.

815

:

Uh, but are, are y'all finding that

the, uh, the GCs and the, and the

816

:

subs have like dedicated iPads and

iOS devices that are there as well?

817

:

What does

818

:

Jesse: Yeah.

819

:

The.

820

:

They do.

821

:

Uh, and, uh, there's obviously like,

uh, there are a lot of iOS devices.

822

:

Like you said, people, you know,

the, uh, a lot of, uh, GCs will

823

:

have a dedicated device for this.

824

:

Um, and you know, we're

just a small company.

825

:

And so I think there was a, It

was a very intentional decision to

826

:

say, we're going to start with one

platform, the platform that we think

827

:

is going to be best suited to do this.

828

:

And we're going to try to aim for

quantity or quality over quantity.

829

:

Right.

830

:

Uh, because we could have opted to build a

Unity app and said like, Hey, look you can

831

:

run it on HoloLens and iOS and Android.

832

:

And there was definitely

some push to do that.

833

:

Um, but you know, the

way I see it, it's like?

834

:

once, once we've got, really good

adoption and we're looking to grow

835

:

past that adoption, then, you know,

it'll be a great, a great decision

836

:

for us to have to make and say like,

yeah, let's expand to other platforms.

837

:

Um, uh, but the, you know, we just have

to decide how to use our resources.

838

:

Randall Stevens | AVAIL: sense.

839

:

Great.

840

:

Evan Troxel: I, I have, I have one more

topic, maybe mini topic to kind of close

841

:

us out, and that is just about technology

adoption on the construction side.

842

:

You guys are bridging the gap

between design and construction here.

843

:

Uh, design intent.

844

:

And that being built to the

original intent is where

845

:

these problems occur, right?

846

:

Like you talked about it in the

very beginning about communication

847

:

problems being the biggest challenge

for rework having to happen, right?

848

:

So, um, I know there's a lot of

contractors out there who are looking

849

:

to minimize the rework, looking to

minimize the, the communication problems.

850

:

So when it comes to adoption and your

story to them, and then maybe the,

851

:

the outcome, if you could just talk

about that from an industry standpoint,

852

:

from your perspective of a, a smaller

tech company startup, who's solving

853

:

real problems on the job site in

between design and construction, like,

854

:

What's the adoption story been like?

855

:

What are you seeing?

856

:

Are you pleased?

857

:

I mean, I'm sure it could always be

better, but, you know, working as a

858

:

digital practice leader in architecture,

like, technology adoption is so,

859

:

so difficult in our industry to get

people to modify their tool set, the

860

:

way they do things, better practices.

861

:

Technology adoption can lead

to a lot of gains, but it takes

862

:

a lot of dedication to do so.

863

:

So I'm just curious from your segment of

the market, what your experience has been.

864

:

Jesse: Um, Yeah.

865

:

that's a good question.

866

:

I, you know, the first thing that

comes to my mind is the, It is

867

:

basically that you have to, that I

think we've put a big focus on really

868

:

listening carefully and trying to, um,

869

:

it, it, it doesn't do anyone any good

to kind of come at this and go like,

870

:

look, we've solved all your problems

and everything's perfect and you

871

:

just need to adopt our technology.

872

:

Right.

873

:

Um, the reality of it is that we're

partnering with people to, we're You know,

874

:

try to get this tech to a place where

it is really valuable for real users.

875

:

And, uh, that means that they experience

pain points using our software.

876

:

Uh, that, and it means that when they

do, we try to approach them with a

877

:

lot of empathy, a lot of curiosity.

878

:

Um, and, uh, and that generally people

respond really well to that, right?

879

:

Like if you're coming in really

interested in their actual problems,

880

:

uh, and you're not coming in trying

to claim to have, you know, solve them

881

:

without, without, you know, uh, you know,

full context, um, that they're willing

882

:

to kind of walk that path with you.

883

:

And it's like you said, it

isn't a quick path, right?

884

:

But these aren't, you know, Easy problems.

885

:

Um, and so I think that we are really

seeing, uh, our software starting

886

:

to help people do their work,

which is what we're trying to do.

887

:

And, and it's getting there because

people are, have been very patient

888

:

with us and willing to try it at

various stages, um, of development

889

:

provide us really honest feedback,

uh, and that we do our best to really

890

:

incrementally, you know, move it along.

891

:

Randall Stevens | AVAIL: That's a

892

:

great tip.

893

:

Oh, go ahead, Adrian.

894

:

Adrian Hatch: I was just going to

say it's, it's treating technology

895

:

as another tool, you know, like,

it's, it's not this other field.

896

:

Uh, that's totally novel to construction.

897

:

Construction has been applying

technology since the industrial

898

:

revolution and even earlier.

899

:

Um, I think the way we've

collectively thought about

900

:

selling technology kind 15 years.

901

:

We've had 20 years really had a big

boom of that in the office side.

902

:

And bringing it to construction

is not the same as bringing it

903

:

to a traditional office setting.

904

:

It's, it's much more about augmenting

that human and their decision

905

:

making, their existing experience

out on site with another sharp

906

:

tool that's going to help them out.

907

:

And that's that empathetic, that user

driven, that product led growth story.

908

:

I think that's the key to success.

909

:

It's gotta be part of your

DNA from the very start.

910

:

You can't start externally and then

come out to the site and say, All

911

:

right.

912

:

now to figure out how

to make this thing work.

913

:

You guys all figure out how to

make this thing work here now.

914

:

Randall Stevens | AVAIL: Yeah, I think I

was going to make a comment that, um, you

915

:

know, I was a, uh, a big, uh, consumer

of the Steve Blank lean, you know, the,

916

:

the whole methodology that came about,

you know, really in the late:

917

:

when Steve Blank started espousing it.

918

:

Yeah.

919

:

You know, we, and we try to follow it.

920

:

It's this whole customer, customer

development methodology of like not

921

:

trying to go, you know, solve this, uh,

you know, in a vacuum and then magically

922

:

bring it, bring it to market and think

that it's going to be, you know, a

923

:

great solution for customers problems.

924

:

But I still find that.

925

:

Even with that mentality from the

development side, a lot of times it's

926

:

hard to find customers who are, who

either, you know, say, just let me know

927

:

when you're done and have it solved.

928

:

Because it does take a lot of back

and forth and their willingness

929

:

to give you their time and input.

930

:

So maybe you can talk about, you

know, how you've identified those

931

:

kinds of customers, what are

good, and early things to ask them

932

:

and how do you get them

933

:

involved in that process and

understanding that, you know, if,

934

:

if you really want this to work,

I need your time and expertise,

935

:

um, to contribute to this process.

936

:

Jesse: it's a big ask.

937

:

I, you know, cause it, and, and just kind

of from a business standpoint, right.

938

:

Uh, you know, like I think we have

a good product right now, but it's

939

:

continuing to grow and, you know, uh, you

basically are asking someone like to be a

940

:

partner, but also ideally pay us because

otherwise we can't keep doing this.

941

:

Right.

942

:

Um, yeah.

943

:

And Yeah.

944

:

I don't know that there's like

a quick, easy answer to that.

945

:

Uh, I know that we've had a good team

that has worked hard to just kind of

946

:

foster personal relationships with

people so that that ask feels more

947

:

reasonable, uh, that they have more

trust that like we will actually, you

948

:

know, incrementally deliver on that so

that they're not just sinking their time

949

:

in and not going to get something out.

950

:

Um, But yeah, I, I, I don't know.

951

:

There's like a shortcut there.

952

:

I think it is a, it puts, it puts a fair

onus on our, uh, you know, on customers.

953

:

I will say, and Adrian

can speak to this too.

954

:

Like, you know, we, we

do work really hard.

955

:

We, we don't want users testing

our software for quality.

956

:

We want qualitative feedback.

957

:

Um, And, uh, you know, that means that

we do put a pretty big emphasis on

958

:

making sure that by the time it makes

it into someone's hands, it, it works.

959

:

It, it may not be the, the

product that they, they want.

960

:

Uh, and you know, there, it's not to say

we'd never ship a bug, uh, but that's

961

:

an area that we care a lot about, um,

and that we're constantly reflecting

962

:

on and trying to understand how to do

better on, um, because I think that

963

:

relationship requires that, you know.

964

:

It's, it's a big ask, requires a lot

of trust, um, and we value it so much.

965

:

And we get a lot out of it beyond the

simple, you know, the, the bill that

966

:

we may send at the end of the month.

967

:

Randall Stevens | AVAIL: Yeah, I

think with some of the larger firms,

968

:

especially where they go back and forth

from doing some in house development

969

:

or have people on their team that

can do some technology development.

970

:

You know, those, those kinds of

companies start to understand.

971

:

It's like, would I rather, do I want to,

do I want to own these lines of code or

972

:

do I want to contribute and have a third

party like yourselves own that, but I'd

973

:

realize it's cheaper for me to, to spend

time and give input, and then I'd rather

974

:

have you maintain that and I'll buy the

product, but it's, but you own the code.

975

:

Right.

976

:

That's

977

:

Jesse: yeah, we do have, you know, we

have strategic partners that are, you

978

:

know, invested in various ways that kind

of bridge that gap where they have, you

979

:

know, uh, Some stake in the company's

general success, but then also, like, you

980

:

say, they just want this functionality.

981

:

Like they know that they want, uh,

to have access to BIM in the field.

982

:

Right.

983

:

And, uh, yeah, that's not

their expertise building apps.

984

:

Right.

985

:

So, uh, yeah.

986

:

Adrian Hatch: providers roles here

is to de risk that new technology

987

:

to make sure that those is essential

issues that, uh, well oiled software

988

:

team can identify in advance.

989

:

Don't make it out to the site.

990

:

It's something that.

991

:

Really took me a while to acclimate

to in this industry is how much of

992

:

a risk adopting a technology is.

993

:

You know, it's always a risk.

994

:

Like, every book on sales for tech

will go over how, you know, there's

995

:

the risk side for your customer.

996

:

But on a job site, that risk can

be really expensive really quick.

997

:

In the worst case scenario, that can be a

risk to a human being's health and safety.

998

:

And so, you know, Making sure that

what you provide is ready for that

999

:

kind of here's how it could work better

feedback, not it didn't do what it was

:

00:54:36,161 --> 00:54:38,541

supposed to, I think is pretty key.

:

00:54:38,551 --> 00:54:42,881

That's where you're doing your part as

that solutions provider by saying, Look,

:

00:54:42,891 --> 00:54:44,671

we We might not have solved it the Right.

:

00:54:44,671 --> 00:54:45,211

way yet.

:

00:54:45,251 --> 00:54:47,431

This might not be your ideal solution.

:

00:54:47,701 --> 00:54:51,391

This might not be the perfect

screwdriver, but it's gonna turn a

:

00:54:51,391 --> 00:54:53,321

screw if you do what we tell you to.

:

00:54:53,321 --> 00:54:57,491

It's not gonna put a hole in the wall

if you, if you press the wrong button

:

00:54:57,491 --> 00:55:02,171

or something like that, and I, I think

that's often been a challenge for

:

00:55:02,231 --> 00:55:07,311

early stage technology adoption and

construction, especially new, cutting edge

:

00:55:07,311 --> 00:55:11,641

technologies like AR, where we haven't

figured out all those pitfalls yet.

:

00:55:12,016 --> 00:55:17,576

As a, as a, you know, tech, uh, field,

there's still places where we haven't

:

00:55:17,586 --> 00:55:22,136

really hammered out all of the, uh, the

potential risks to a development team.

:

00:55:22,696 --> 00:55:26,596

Making sure that those risks don't make

it out to the job site, that they are at

:

00:55:26,596 --> 00:55:31,866

least captured on the, on the development

side, uh, prior to ever being something

:

00:55:31,926 --> 00:55:33,486

that can cause an issue out in the field.

:

00:55:34,399 --> 00:55:34,819

Randall Stevens | AVAIL: Great.

:

00:55:36,039 --> 00:55:40,579

Well, I think, uh, we're going

to want to have you guys back as

:

00:55:40,609 --> 00:55:43,449

that feature, uh, evolves, right?

:

00:55:43,449 --> 00:55:46,399

So, I think we've already talked

about that, that maybe over the coming

:

00:55:46,399 --> 00:55:47,389

months we could have you guys back.

:

00:55:47,669 --> 00:55:52,219

back, back on and kind of showing a

little more of the behind the scenes

:

00:55:52,219 --> 00:55:56,019

development of this and how this is

evolving and working in the field.

:

00:55:56,059 --> 00:56:00,489

But, uh, we definitely appreciate you

all coming on and sharing, you know,

:

00:56:00,579 --> 00:56:02,399

some of the thinking and process.

:

00:56:02,459 --> 00:56:07,499

Um, like I said, I think, uh, I'm

interested longer term in like the

:

00:56:07,499 --> 00:56:11,389

place where we sit and are doing work

with our customers is like, is there

:

00:56:11,399 --> 00:56:15,579

something that could be being done with

the data that we're helping to form?

:

00:56:16,044 --> 00:56:20,104

Earlier in the process of the downstream,

it's better suited for the kinds of

:

00:56:20,104 --> 00:56:21,844

things that are happening in the field.

:

00:56:21,844 --> 00:56:25,424

And where's the handoff of this

data and those types of things.

:

00:56:25,424 --> 00:56:29,514

So for sure, we're going to want to,

uh, I'm going to want to take a special

:

00:56:29,514 --> 00:56:32,234

interest in kind of keeping track

of what you are doing on this front.

:

00:56:33,514 --> 00:56:34,524

Jesse: Yeah, that sounds great.

Follow

Links

Chapters

Video

More from YouTube