iPhone 17 Pro LiDAR vs. Survey Total Station Accuracy

What is the mapping accuracy of the

iPhone 17 Pro's camera and LAR sensor in

comparison to a high accuracy surveying

total station? Okay, so I'm going to

start on this corner and we're going to

start collecting data. And I'm going to

slowly move around the building. And as

you can see, the points are showing at

the bottom. These are our LAR points

that are being generated. And today I'm

actually using the Pix 4D catch app to

collect data. You can download it for

free in the app store. And that's

because what you're looking at are the

points being generated from the LAR

sensor into our point cloud. So if you

look here, this entire model is going to

be reconstructed in real time thanks to

the LAR sensor and that's going to give

us the data we need in order to do

surveying with our iPhone. In the final

stretch on the last side of the

building, you can see we've got a water

fountain here. And I want to actually go

inside and get beneath the canopy in

this seating area. So you can see this

area nice and clear. We've also got the

doors to the bathrooms and the other

side of the seating area. And that will

complete our scan. And so here we can

see our entire scan project. We've got

all sides of the building and plenty of

detail here under the canopy. And in a

matter of just a few minutes, we were

able to generate an entire 3D model

using our iPhone 17 Pro. Now, let's do

this traditionally using our total

station. So, to start, we're going to

set our first control point here on this

side of the building.

And the second control point I'm going

to set is going to be right here. It's

got a clear line of sight to the first

control point where our total station

is, as well as being able to see the

back of the building. Total stations

heavily depend on having a control

network in order to operate properly. By

establishing one known point and

backsighting a reference line known as

our backsight and any change in the

angle and distance for our foresight

readings will result in the coordinates

of the data that we collect. And using

our simple distance formula and applying

a change of azmouth to the back site

gives us the positions of our new

points. And that's why we're using the

total station as the benchmark standard

to test the accuracy of the iPhone 17

Pro. All right, there we go. Okay, so

now I'm going to show you how we set up

our total station.

So, here's our point right here.

And you can put the total station right

on top of the tripod. There we go. Nice

and tight. All right. At this point, I

like to turn on my total station. Some

total stations will have a lens that you

can look through in order to see the

point below you. But this particular

model actually has a laser that will

shoot down so we can see where the laser

is and ensure it is over the point. All

right. Now, my laser is directly over

the point so I can step on the legs to

secure the tripod to the ground. This is

the first bubble that we need to ensure

is in the center to level out our total

station. And the way we do this is by

adjusting the legs. And now my digital

bubble is all I have left to level. And

I do that by using the screws here on

the side to fine-tune the level of my

total station. And so if I take a look

here, I'm only off by a couple of

seconds and I'm directly over the point.

And that is exactly how you set up a

total station. All right. Now, let's set

up our job so that we can start

surveying using our total station. Okay.

So, I'm going to start by creating a new

job. We'll call it iPhone 17 Pro. And

we'll store it. And we'll create a new

point. This right here will be point

number one. And I'm going to assume a

coordinate of 10,000 in the easting,

5,000 in the noring, and 100 in the

elevation. Okay, I'll say store. And now

we can go back. We'll switch over to

apps, and I'll go to setup. We're going

to set our orientation. All right, this

is the job iPhone 17 Pro. This is point

number one. And what this will do is

it'll measure from the point up to where

the scope of our total station is. So

that came out to 5.056 ft. I'm going to

say okay. We'll say okay again. And now

it's time to define our back site. All

right. So, what I've got here is the

prism as well as Leica's AP20 autopole.

This is an IMU that is going to give us

the ability to hold the rod in any

position. So, we don't have to have it

plum in the center like we traditionally

had to. So, I'll start by powering on

the AP20.

We're going to name this point number

two. Our target height is set to 5.2 ft.

Everything looks good here. So, now I'm

going to have the total station power

search and find us.

lock to target.

>> Okay, perfect. So, now we are locked to

target. Okay, so we're going to hold our

back sight point, point number two. Now

that we've established our back site, we

can begin collecting data. All right, so

the first thing I want to do is raise my

rod height to like 6 ft. So, it's above

my head. The height automatically

updates. And then I want to

>> initialize my IMU. And it was pretty

quick when it did that. So, I don't have

to worry about plumbing the rod every

time I want to take a shot. We're going

to switch our point number here. Switch

over to 101. And we'll start with

concrete here. And let this be the

beginning of a line. There's our first

point. Come down over here. Next point.

Come over here. Looks like the elevation

here starts to change. So, I'm going to

shoot another point right here. And

we'll call this B L DG for building.

Again, we'll begin a line. Make sure our

IMU is initialized. There we go. And

we're going to store. Nice. We've got an

existing finished floor elevation right

here. So, we'll call this EFF store. Got

another existing finished floor right

here. Measure point stored. And where I

took shots with the concrete, I'm going

to shoot those in here, too. Building.

We'll do another building shot here. Do

another building here. Come over on this

side and take our concrete shot. So,

taking a look here, I can kind of see

what my project is starting to look

like. We've got one line for the

concrete, one line for the building, and

a couple of extra shots for the existing

finished floor. We'll keep it going.

We've got a concrete shot here,

and we'll take our last concrete shot

here. And you can see it here when I

tilt my rod back and forth. The

compensation is calculating the exact

position of the bottom of the pole.

Store

>> here is part of the canopy. So, I'm

actually going to create a new point

here.

I'm going to call it can P. Switch it to

begin a line. All right. So, we're going

to measure this corner of the canopy.

I'm going to come over here to the other

side and then we'll measure over here.

Okay. Back over here. Now, I want to get

the concrete corner here and then over

there. And then we want to get like this

part of the wall as well. So, I'm going

to get this corner here of the concrete.

The last thing I really want to do is

survey the inside here. Uh, I want to

get this wall. I want to show what it

looks like. So, I'm actually going to

create a new code here. I'm going to

call it wall. We'll begin a line.

>> And make sure the total station finds

me.

>> Get this corner here. Call this wall

one. Okay. Come over here. Call this

wall one.

>> We're going to shoot here. We'll shoot

this corner.

Come back over here. Shoot this. I'm not

exactly sure I got it all. So, I'm going

to double check by zooming in. Actually,

that looks pretty good. So, we got this

whole L shape right here. We got to do

the other one on the other side now.

All right. So, I think I'm going to set

the third control point here. Important

that we have visual line of sight to our

total station so that when we set up our

total station here, we have a known

position.

Okay,

there we go. Okay, and I'm going to hold

this point and we're going to call this

point number three. All right, so now

we've got point number three here. All

right, now let's pack up the total

station and set up on point number

three. All right, so we've got the total

station now set up. We're going to

assign it to point number three. And the

height, we can measure it. We have

5.078.

Okay,

as for the back site, we're going to be

back sighting the point we were just set

up on. So, this is point number one. All

right. And now we can go into the

measure menu. And there's the total

station set up on point number three.

All right. Now, let's continue surveying

this side of the building. Now, we

stopped on this corner of the concrete.

So, I'm going to continue here on this

corner. We're going to start our tilt

compensation. We're going to initialize

our IMU.

There we go. We're going to store the

point.

We'll finish up the building and then

let's get an existing finished floor

elevation on this door. There it is.

Store.

>> All right. Love that. The building is

going to keep going that way.

Measure. And I can't see this back end

of the wall, but that's okay cuz we've

got another setup that we'll do so that

we can see all that information. So if

you find surveying and geospatial

technology interesting and you want to

connect more in the industry, then I

invite you on February 16th to the 18th

in Denver, Colorado to attend Geo Weeek

2026. Geo Week is the largest surveying

and geospatial technology conference in

the country with over 4,000 attendees

and hundreds of exhibitors from all over

the world. I'm going to be there along

with many of our students from the

survey school connecting and networking

in person as well as attending different

conference sessions and learning as much

as we can. For a limited time, you can

use promo code ramy 100 to get a $100

off of your full conference pass or your

exhibit hall pass. Check out the link in

the description and I look forward to

seeing all of you in Denver for GEO week

2026. And now, just like last time,

we're going to set one more control

point. And it's important that this

control point is one visible from our

total station, but also visible from

control point number two. It's important

that we're able to see that control

point because we need to be able to

close our traverse. Closing our traverse

will allow us to run any kind of

adjustments if we have any major error

and give us more control over our

control survey. Let's set the last point

here.

All right, looking good. All right. Now

that we got this measurement here, we're

going to bring the total station and set

it up on point number four.

Okay, perfect. Now I'm going to shoot a

back sight over to point number three.

Okay, set. And now we'll shoot a second

back sight to point number two.

And set. We've got this back corner of

the wall, this corner of wall four, this

corner here, finishing up the corner

where we started. And so the last thing

I need to do is use the Leica GS18 GNSS

receiver to get geodetic positions on

all of our control points. Having

geodetic coordinates on our project will

allow us to transform the entire project

to stapling coordinates and allow us to

analyze the absolute accuracy between

the total station and the iPhone 17 Pro.

So, I've gone into NGS's website and

plugged in our location. And using their

coordinate conversion and transformation

tool, I can see that the scale factor

here is for grid to ground. So, I'm

going to copy this number. And now I can

apply this scale factor to our project

and do a proper transformation for our

coordinates. Now, one way to improve the

accuracy of the iPhone 17 Pro is to

introduce an RTK GNSS antenna. Now, Pix

4D works with a lot of different brands,

but this is the one that we're going to

be using today. This is the Bad Elf Flex

Mini RTK GNSS receiver. It provides

centimeter level accuracy and integrates

with Pix 4D catch and your iPhone 17

Pro. So, the way this kit works is you

get a handle like this. You attach the

mini right here. It just screws in and

you take your iPhone

and then that turns just like this. And

now you've got an RTK enabled GNSS

receiver sending high accuracy

positioning to your iPhone while you're

doing your mapping. So let's do the

survey one more time using our iPhone 17

Pro while it's attached to the Bad Elf

Mini. All right. So as you can see here,

I've got a fixed reading on Pix 4D Catch

and we are going to start collecting

data just like before. It's exactly the

same except now we're getting RTK

corrections to all of the positions of

the iPhone. Coming around here and like

before you can see the entire building

coming to life in real time and that is

thanks to the LAR sensor on the iPhone.

Coming around this part of the building,

we'll get the back side and we'll get

this area here. We'll finalize this. As

expected, I did lose my RTK, but that's

okay. All right, and let's come on out

from underneath the canopy. And taking a

look here, I can see all of the details

that I need for the building. I can also

see all of the concrete that we also

measured with our total station. So,

it'll be really good to be able to

compare the point cloud to the data

collected with the total station. All

right, now that we finished collecting

data with just the iPhone 17 Pro, the

Total Station, and the iPhone with the

Bad Elf Mini, let's head inside so that

we can process all of these data sets

and see our results. All right, now

let's go ahead and take a look at all

the data. Now, inside of Pix 4D Catch,

you have the option to upload all of

your data to Pix 4D Cloud by selecting

upload and clicking next. You can see I

have a lot of different settings that I

can go through. The first is a region of

interest. So, if I want to crop out

certain parts of my point cloud, I can

do that. I also have the option to add

in Gausian splatting. This will give you

an immersive 3D deliverable that you can

use in order to fly through your site

and provide the perspective as if you

were on site and looking around. This is

really great for 3D deliverables for

your clients, and it's only available on

Pix 4D Cloud. You have some of the more

traditional outputs like volumes and 2D

maps, and you have the ability to add in

your coordinate system. Now, if you're

looking for a quick deliverable that you

want to just share the link to your

client with, then Pix4D Cloud is going

to be your best option. But if you're

looking for a survey grade model that

you want to be able to work with and

have full control of, then I recommend

that you use Pix 4D Matic. So, if you're

looking for that, you're going to click

on this little box and arrow at the top

and select export all data Pix 4D Matic.

And this will start to export your data

for you. And you can now upload this to

Google Drive and open it up on your

computer. So, as you can see here, I've

got two different data sets. I've got my

single GNSS data set that had just the

standard GPS on my phone, and we have

the RTK data set that we collected with

the help of the Bad Elf Flex Mini. Now,

Pix 4D just released Pix 4D Matic Pro.

This is their 2.0 version of Madic,

which combines traditional Pix 4D Matic

with Pix 4D Survey. And I'm going to

show you exactly how to navigate this

software. And be sure to stick around

until the end of the video because I'm

going to show you how you can get a free

Pix 4D Matic Pro license in order to

process your iPhone and drone data. All

right, so we're going to load up Pix 44D

Matic here and I'm going to give this a

job name. So, let's call it Dodge Park

Bathrooms. Start. And then all we need

to do is just drag and drop our images.

So, I can just take our images here and

drop them in place. And as you can see

on the main screen, I can see our

trajectory information for our iPhone as

we walked around the building. We also

see all of the images that were

collected, which is really nice. And

regardless if you're just using the

phone or you have an RTK GNSS receiver,

the process is exactly the same. You

just drag and drop it and everything is

imported from within your project

folder. Now, we can come down here on

the pencil and actually update our

coordinate system. So I'm going to be

using NAD83 Michigan South and I'll be

using NAVD88

with goid 18. So I will apply. And now

we've updated our project coordinate

system. And you can see the map is

loading up. And there we go. That is the

bathroom. And we circled around it.

Everything looks good. So in the

processing menu, we're going to be

selecting calibration. This will allow

us to process all the data between the

LAR sensor, the camera sensor, and the

GNSS receiver. You have the dense point

cloud which comes out of the camera

sensor. And I'm actually going to select

both because I want the fusion point

cloud which combines both the camera and

LAR sensor. You can also process out a

mesh, a DSM, an ortho image. I like to

wait on those. I like to actually get my

point cloud first and then I can process

out that information. And it's as simple

as that. Now I'm just going to click

start. And depending on my computer

hardware and how many images I'm working

with, this could take a few minutes up

to a few hours. So, we'll come back once

we've processed these data sets and see

how they look. All right. So, the first

data set we're going to look at is just

the iPhone with the camera and LAR

sensor. This is what an $1,100 cell

phone can do when it comes to doing 3D

mapping. Taking a look here, you can see

this is pretty incredible. Just from the

camera and LAR sensor, it's incredible

to see an entire 3D model of this

building being generated. Decent amount

of noise here. It's not too bad. nothing

that software couldn't clean up. Now,

coming on the back here, I do see some

misalignment here. You can see this is

part of the wall, but then it's like

shifted down here. There is a little bit

of an issue. Yeah, you can clearly see

we have some issues with alignment here,

and that's because single position GNSS

only has an accuracy of about 3 to 5 m.

So, our trajectory information isn't

necessarily great. We are heavily

depending on the camera and LAR sensor

to help us with the alignment of the

data using methods like structure from

motion in order to build out our point

cloud. But nonetheless, this is a pretty

nice model and we will be looking at the

accuracy of this model in just a few

minutes in comparison to our total

station. Now let's see what the 3D model

looks like having added an RTK GNSS

receiver. And so this here is the RTK

model. You can see we have closer

estimation between our initial camera

position and where it was calculated.

And that's because we have high accuracy

trajectories on our phone's GNSS. And as

a result, that means there's less guess

work and a cleaner data set. So this is

the exact same settings. I didn't add

any filters here. You can see we have a

lot less noise than before. Still some

noise. I mean, it is a cell phone after

all, but it did a pretty decent job of

reconstructing everything. And in the

back here, we don't have that same

alignment issue with the wall that we

had with the other data set. So, that's

good to see. Now, I want to show you

what the data that came out of the total

station looks like and what we want to

achieve when it comes to creating

deliverables for our clients. So, this

is what the total station data looks

like. And while it's not a sexy point

cloud, there is a lot of valuable

information here that we're going to be

using as our benchmark data set to

analyze the accuracy of the iPhone 17

Pro. And so what we're going to do is

use the survey functionalities in Pix 4D

MATIC in order to extract information

from our point clouds and compare it to

what we have from our total station. So

if I come over here to the surveying

tab, I can see I have a bunch of

different options here. We'll start with

terrain classification. And what this

will do is actually classify the ground

versus the building. So I've got my

settings here. I'm going to click

classify terrain. And there we go. In

purple is the building and in yellow is

the ground. I do have the option to

manually fix this. So, if I've got areas

that I know for sure are the ground that

weren't classified correctly, I can just

select them and switch them over to

ground. So, there we go. Now, this is

part of the terrain. Apply here. And

yeah, that looks pretty good. It doesn't

have to be perfect. It just needs to

capture as much of the terrain as

possible. Next, we're going to do grid

of points. This is going to allow us to

create a digital elevation model or DEM.

A DEM is going to be a scatter of points

along our terrain, giving us kind of a

simplified version of our point cloud.

So rather than working with millions of

points, we could work with a couple of

hundred or a couple thousand points. So

I like to do a 3-FFT spacing between my

points, which is about 1 meter. The Z

range, I like to keep this somewhere

between 3 and four. So we'll keep it

here. And maximum number of points, we

can just keep this set to 2,00. Okay,

let's generate our grid. And there we

go. We've generated our DEMs. I don't

know why I have points up here on the

tree. That's okay. I can just select

them. I can just come over here to my

selection tool and I'm going to say

select only grid points and I'll select

them and delete them. Make sure I don't

have any other outliers like that. I

think we don't necessarily need this

data over here. So, I'll delete this

stuff. Same with over there. Okay. I

think this is pretty good. Yeah. Okay.

I'm happy with this. Now we can go into

our triangular irregular network or tin

which will connect all of these points

together and generate our surface. So

there we go. We've generated our surface

from those points. And actually I'm

going to turn off the point cloud so you

can see what this looks like. And this

is basically the ground below the

building how it is laid out to get a

better idea of what this looks like. I

can actually generate contour lines to

help us identify where the increase or

decrease in elevation is. There's also

some object detection features here like

manholes and poles, but that's not the

type of survey we're doing. If you're

doing a topographic survey, then you can

use some AI functionality to extract

that information from your data set.

Now, the one other thing that we can do

is actually extract features in the

point cloud like our buildings, our

sidewalks, and the walls so that we can

mimic exactly what we got with our total

station. So, what we're going to do is

come over here to layers, and we're

going to add a new layer. And we'll

start by calling this sidewalk. And I'm

going to switch the color here to cyan.

So now I can select a polyline and

simply just click on the points for

where the sidewalk is. So if I have a

break there, I can stop. I can start

over here and continue. And yeah, this

right here is what we call feature

extraction. So if you make a mistake,

you can just simply hit undo and fix

your mistake. Okay, there we go. And

then if I want to start from a point, I

can just select that verticy and work

off of it. And it's basically just like

tracing. I mean, you're tracing out the

points as they appear in the point

cloud. Definitely can see some noise

there. Yep. So, you got to be careful.

You're not picking the noise and you're

actually picking the actual ground. Keep

selecting points. Coming up here, we can

pick the corner and then pick this and

we'll just end it right here. Perfect.

So, now we've extracted the sidewalks.

Next, I want to do the wall. And I'm

going to switch the color to maybe a

little bit more green. Yeah, like that.

Yeah, that's good. So, there's wall. Got

a wall here. Selecting the same corner

there. It looks like the building line

is right in line with this grout line.

So, I will just select the grout line.

Call that good. Okay. Next wall. Same

thing. It looks like it's right with the

grout line. So, we'll just select right

here. And what we're doing essentially

is just like what we were doing in the

field. We are doing a topo. We are going

around and selecting the points that

belong to those features. So, it is very

very similar to an actual too survey,

but it's like virtual. You're doing it

on your computer in the comfort of your

office with less environmental error

because you've already captured all of

the data using geospatial sensors. The

last one I want to do is the building.

We'll give this a little bit more of a

brighter yellow. Ah, this yellow's fine.

And we'll select every two sidewalk

blocks and here. And it looks like

okay. So now that we've got the

buildings, the walls, and the sidewalks,

we're going to add in just single

points. We're going to add in the

existing finished floors of the doorways

that we had measured in with our total

station. So we're going to click add

existing finish floors. And let's change

this color to purple. Okay. So I'll

switch over to just a marker. And then

here we got one existing finished floor.

I'll just select it. Come over and

around. Here we've got another one right

there, one right there. And we actually

have two more that we missed with the

total station, which are the bathrooms.

So, one right there, and one right

there. Perfect. And then the last layer

I'm going to add in is our canopy. And

let's make this one red. And really, the

canopy is just going to connect between

these two. So, I'll switch over to

polyline, and we'll just connect between

the two walls. And there we go. Now,

we've got the canopy above us. Just like

that. And so to improve the tin, what

I'm actually going to do is identify

which of these are terrain layers.

Actually, all of them are terrain

layers. So, we can turn them all on. And

these are just points that we are

drawing on the terrain, which we are.

And we'll go back to the survey tools.

And now we'll actually enable the use

terrain layers as brake lines. This will

give us a more accurate tin since we're

actually creating brake lines for where

there's sidewalks or where there's a

building, uh, actual structures that are

affecting the ground. and I'll say

generate tin. And there we go. We've got

an updated tin. And we can even update

our contour lines. All right. So now

I've exported both data sets from Pix

4Dmatic. And let's compare them to the

total station benchmark data on AutoCAD

Civil 3D. We're going to do three

different tests. The first is the ortho

image, which is the 2D aerial

perspective of our data. The second is

the 10 model. So this is the surface

model that we generated from our DEM.

And the third is going to be the actual

features that we extracted. We'll do a

point-to-point comparison, see the

difference in X, Y, and Z, and see what

the overall accuracy of the iPhone 17

Pro's camera and LAR sensors and what

the accuracy is when we add in an RTK

GNSS receiver. So, this right here is

the total station data. And what I'm

going to do is do map insert to bring in

the ortho image of the iPhone data with

the RTK. So, this will just give us a

preliminary view of what the

reconstructed ortho image looks like and

how accurate it is to the total station

data. So, here we go. We can see we've

got the sidewalks here and that's really

all we're going to be able to see with

the design of the building. We're not

going to see the building corners. We're

not going to see the walls or the

canopy. So, we can't really judge

anything here. It's just completely

completely hidden from our view. But

what we can see is the sidewalk. So,

let's see how close we were. So, here on

the east side of the site, it looks like

it's pretty close. It's not terribly

off. It does run along the sidewalk

pretty accurately. I do like that. And

yeah, looks like it's just slightly

shifted here, but it's not terrible. So,

again, this is just the ortho image.

This is not going to be the most

accurate deliverable that you can get.

But it's really cool to see that just by

holding your phone like this and taking

some pictures from a, you know, oblique

perspective, Pix 4D has the ability to

generate an ortho image for you. So,

this looks like it came straight out of

a drone, which is really, really cool.

Now, this is iPhone with RTK. If we

would have just done the iPhone data

set, it actually didn't produce a usable

product. So, I'm not even going to show

it to you guys cuz it just didn't work

out. So, if you want this ortho

deliverable, you're going to want to use

RTK with your iPhone. It just won't work

well without it. All right. Now, let's

take a look at the surface models

generated from both iPhone data sets.

What I need to do first is create a

surface of the total station. So I'm

just going to call this total station

surface. Say okay. Okay. And so now if I

go to object viewer, we can see this is

the surface from our total station. All

right. So now we've inserted both

surfaces RTK and standard GNSS. So if I

let's say I want to view all three of

these surfaces together, I see ooh one

is much lower than the other two. Uh, if

I were to take a guess, I'm going to say

that probably the GNSS one is lower.

Let's look at this a different way. We

can see that the tin is much lower for

this surface and this is the GNSS

surface. So, so let's take a distance

between these two. So, from here to here

and looks like our delta in elevation is

a - 5.58

ft. So, just under 2 m below where it

should be. So interesting to see. It's

pretty consistent. If we take a look at

the west side here, yeah, we've got -5

ft. So yeah, it is pretty consistent,

but again, that's to expect with just

single solution GNSS. Okay, now the

ultimate test. Let's see how close the

line work is from the feature

extractions that we did in Pix 4DAT.

We'll start with the single GNSS

observations. Insert. Okay, there we go.

And yeah, I mean it looks like well

there's definitely some differences, but

it's kind of cool to see that they look

very similar. So that that is nice.

Okay, so let's use our annotation here.

And we'll start with like this corner.

Looks like we've got a difference of

about 2 1/2 ft in the horizontal and

vertically. I'm not going to check all

the points vertically, but you can see

here in this example, it's 4.66

ft. So we said, you know, they were

about 5 ft off. So that makes sense.

Between these two points, we got about

2.26 feet, 2.3 feet, 2.5 ft. So yeah,

these are all like right around a meter,

maybe a little bit less than a meter.

This one is 3.88 ft. So this is a little

bit over a meter. The the rotation here

is really an issue. Like I'm noticing

the angle is completely different. It

does kind of align again because

probably, you know, Pix 40 catch will

bring everything together, but yeah, I

mean clearly there's misalignment

issues. Um major differences here. Yeah,

3.4 4 ft. So, I'm seeing differences of,

you know, between two and 4 feet, let's

say. Again, for single GNSS, that makes

sense. That's why we have this. That's

why we have RTK on our iPhone. So, let's

take a look and see how accurate the

iPhone 17 Pro is with an RTK receiver

like the Bad Elf Flex Mini. My god, you

can barely tell the difference. They are

really, really close. I love seeing

this. We'll start here in the east. a

difference of 2/10 of a foot. So that's

about 6 centimeters. And elevation wise,

we're at 0.04

feet. So 4 hundredths of a foot. We're

at like 1 cm there. So that's really,

really good. Check over here. Yeah,

we're within a tenth. Down over here,

about 11 h00s of a foot. So 3 cm between

here and here. 500s of a foot. So like 1

and 1/2 cm up here, maybe a little bit

more. Yeah, about 2/10. So right around

6 7 cm. This corner here, we're about a

tenth. So 3 cm. And then let's look

inside here. Like this is really, really

close. We're at 5 h00s of a foot. This

one's about a tenth. I'm not sure what

happened here, but still pretty close.

Yeah, about a tenth here. And looking at

800s here. And this one's at 700s. About

900s. This corner right here looks

really close. Yeah, 500ths of a foot. So

we're seeing pretty consistently, you

know, a lot of 1 and 1/2 cm. So that is

really nice to see. The corner here is

about a tenth. Here we got 17 hundreds.

And then in this back corner, the back

corner of the building were about a

tenth. The walls here maybe a little bit

more. 2/10th. I mean, that could be just

me picking the wrong point. It's a

little bit more difficult to do that.

Yeah, here it's a tenth. And for full

transparency, something I didn't

necessarily do is actually go and look

at the images. So, when you set a point

like this, you should go through each of

these images and see like you see I'm a

little bit off here. So, I should make

an adjustment to make sure that I'm

really getting that bottom part of the

corner right there. Yeah. So, I can go

into these images and get really, really

precise, which you could see right

there. I could tell I was off slightly.

So, keep that in mind. Also, when it

comes to corners like this. So, how

accurate is the iPhone 17 Pro? By

itself, you're looking at about a meter

to 2 m of accuracy or 2 to 5 ft. And if

you're using a GNSS receiver like the

Bad Elf Flex Mini, then you can expect

an accuracy level of about 2 to 9 cm or

about 1 to 3/10. Special thanks to Pix

4D and Bad Elf for sponsoring this

video. And if you're looking to get a

free Pix 4D Pro license, then consider

joining the survey school. Inside of the

school, we actually have a partnership

with Pix 4D that gives our students a

free 12month license. So while they're

learning about surveying, they can

actually get some hands-on experience

with industry software. So if you're

interested in learning more about

surveying and geospatial technology,

then consider subscribing to the YouTube

channel and visit the surveyschool.com

if you want to elevate your survey

knowledge. Thanks guys for watching and

I'll see you all next