Elevate crash scene documentation with Skydio and Faro
Drones are revolutionizing the way law enforcement gathers and analyzes critical evidence in crash and crime scene investigations. The unparalleled aerial perspectives allow for rapid, comprehensive scene documentation. Skydio X10 enables onboard generation of orthomosaic images and 3D modeling, so you can clear a scene with the confidence that you have all the data you need. Industry experts from Skydio and FARO share their technology solutions for public safety, designed to enhance scene documentation and create compelling courtroom-ready deliverables.
Learn more about crime and accident scene documentation with Skydio.
- So incredibly excited and happy to be
- reunited and joined by my good friend, mentor, and crash reconstruction
- expert, Russ Bloomington. Russ, go ahead and take a moment to introduce
- yourself and tell everyone about your extremely accomplished background.
- Yeah. Good morning or good afternoon, everyone.
- Uh, thank you Noreen and Skydio for having me.
- Um, and it's great to work with you again, uh, too, Noreen.
- Um, uh, my background is, uh, former law enforcement, retired from New
- Hampshire and, uh,
- a former Marine. Uh, spent some time in Afghanistan.
- I'm a, a crash reconstruction expert.
- Uh, and, uh, I've testified as an expert in courts in
- most of the New
- England states. Um, currently, uh, with you from, uh,
- from the Denver, Colorado area. Um, uh, so looking forward
- to today and, um, uh, thank you everyone for joining us.
- Awesome. Thanks, Russ. I'm very humbled too.
- I feel like you missed out a lot of very important pieces.
- [laughs]
- Yeah.
- Um, my name is Noreen Charlton, and I lead strategy
- and marketing for public
- safety here at Skydio. But most importantly,
- I spent more than a decade
- as a crime scene analyst with the Las Vegas Metropolitan Police Department,
- as well as several years with FARO Technologies dedicated to
- crash and crime scene documentation, reconstruction, and forensic
- analysis. So this topic has been my life for
- many years now, and I'm coming to you from just outside
- Cleveland,
- Ohio. So to any of my Ohio friends logged on,
- you can reach out to me at any
- time.
- Whether you are using traditional methods for scene measurements,
- anything from hand measurements to total stations, or you are using
- 3D laser scanners, drone data capture can be integrated
- into your current workflows or replace some of those
- existing methods.
- Many agencies have completed studies on the use of drones
- for speeding up their time of measurement on scene, and on
- average, they're finding that drones are 75% faster than
- some other methods.
- Additionally, from a crash perspective, the h- Federal Highway
- Administration, uh, states that every single minute that an
- accident continues to be a hazard, it increases the, uh,
- chance of additional accidents by almost 3%.
- And for all of us that have worked those scenes,
- I don't know that I can remember
- any fatal accident that I was out on that a car didn't decide it
- just needed to drive through the tape and the cones, and that's the way that it
- needed to go home. Clearing roadways fast is imperative.
- Of course, drones aren't ideal for all situations.
- They aren't the best choice for mapping evidence on the interior of a
- residence, but they are an excellent tool for all outdoor
- scenes from shooting incident that covers a long stretch of roadway
- to a use of force or to a major accident or fatal
- crash, which is what we are focusing on today.
- What's most important, other than using drones just for image
- capture, is understanding what you can do with
- those images
- and how you can take those individual images captured by the drone
- and create compelling courtroom presentations to assist you
- in telling the story in court to the triers of truth.
- Drones don't just provide additional photographs.
- I can't tell you how many conversations I've had with people that are like,
- "Great,
- it's another camera, it just flies in the air." Those
- photographs can provide
- you a lot of additional information that you can use in forensic software
- for deliverables.
- Um, first of all, listed on this slide here, uh, those
- images can be combined into one single image, an
- orthomosaic, and that can be used as the foundation
- for your line diagrams, which we all know is the most commonly
- used
- and always requested at courtroom deliverable.
- I'm convinced that no matter how advanced we get with technology in this
- field, we will always be required to have
- that bird's-eye line
- diagram for court.
- That orthomosaic image can be imported into diagramming
- software like FaroZone, and it basically serves as an updated
- at the time of your incident satellite image.
- So instead of having to import a Google or a Bing satellite
- image to assist with your diagram, which isn't an
- accurate representation of your scene at the time, you can import an
- ortho from a drone.
- Next, especially for all of those of you who are
- familiar with the 3D worlds, those exact same images
- can be used in the photogrammetry engine to create a point
- cloud similar to what you were capturing with any type of LIDAR
- or 3D
- laser scanning.
- Once you have that 3D model via the photogrammetry, now we can provide
- perspectives and otherwi- and views that are otherwise difficult to
- understand with just standard photography and certainly
- difficult to understand with these 2D line drawings.
- Having been in the 3D space now for several years, I cannot
- emphasize enough the importance of 3D data in the courtroom.
- Testimony has gone from 2D images and, like,
- horribly painful descriptions of where you're located in a particular
- image in a scene to giving a jury a full
- understanding spatially of the environment as you were there at that
- time.
- And the best part of it all is that these images
- and these 3D point clouds
- can be used in forensic software, specifically FaroZone 3D
- Expert, so that you can create your deliverables
- and all of your
- forensic analysis all in one place.
- You no longer have to move from the drone to one piece of
- software to another piece of software and finally to your diagram.
- You can do it in one place.
- [clears throat] Dang
- it, my slide. There we go. Um, so the first thing
- that I wanna talk about that
- is a feature kind of of Skydio drones is our Map Capture.
- We've already spoken to 2D diagrams and how important they are for you.
- And again, this is taking all of those individual
- images captured by a drone and
- making them into one single image, an orthomosaic.
- With our latest Skydio X10 vehicle, once you have completed a
- map capture, you can export that image, that
- orthomosaic image, out as a TIFF. And again, that can be
- imported directly into the diagramming software.
- So
- on board my drone, I can create my TIFF.
- I can export that out and bring that right into Fero Zone
- and start my line
- diagram. And that's kind of what you're seeing here
- to the right.
- So on the top, just the orthomosaic image, and on the bottom you can
- see that I just pulled out some quick dimensions, added some white lines there
- for my curbs. And in Fero Zone, I can just turn that image
- off and have my line diagram and be done for my diagram for
- court.
- Here's another example of 2D map capture, and Russ will speak more
- specifically to this case here in a little bit.
- But again, at the top you're seeing the orthomosaic
- image.
- So again, that's all of those images stitched
- together from the drone.
- And then he just traced over it in Fero Zone to get a 2D
- line diagram, and that's what you're seeing on the
- bottom.
- So he just turned off the image and presented just the line diagram.
- So another example, um, of something that you can do with a Skydio drone is a
- software that we call 3D Scan. And in a second here, I'm going to
- play a video that shows you the process of capturing that
- information.
- Um, what you're seeing here is on the drone setting up a
- 3D scan. So right now it's looking at its
- environment.
- Once it's done, you're gonna set some pillars
- and some volume.
- I'm gonna sped this up a little bit just in the interest of time.
- So we're setting up our pillars around the, uh, vehicles
- here, and you'll see that we're gonna set some volume,
- so a ceiling and a floor.
- Once you've done that, you can set some other, uh,
- settings there on the
- drone, and the drone just goes off and does its thing.
- So you can see that it's capturing now it says six hundred and ninety-seven
- photos.
- Once you've set that volume, once you've set those pillars,
- it knows what it needs
- to do, and it just goes off and d- and does it.
- So you don't have to worry about controlling it.
- You don't have to do anything manually.
- It's taking all the images that you need with the intention of
- making a 3D point cloud at the end, capturing all of
- the side lap and overlap in different angles you need with the
- intent of creating that 3D point cloud.
- You can see here in the video it's almost done, close to that.
- Um, and then it just ends there, and you can create the onboard model,
- which is what it's doing, processing that scan.
- And it will create an onboard model and then of course export that map capture
- for you as well. We're gonna talk specifically about what
- an onboard model is and
- what that looks like.
- So specific to 3D scan on our X10, you can now
- generate a model of your scene right on the
- drone, um, before you clear. And so the idea is that this
- ensures on-site confidence that you've captured absolutely
- everything that you need before clearing the scene.
- That model will export as a .gltf,
- and I've worked very closely with the Zone team in the last couple weeks.
- You can now import those directly into Zone as a 3D model
- as well.
- I would say for 3D models in, in crash and crime, we're not, probably not gonna
- be using a mesh model a whole lot for reconstruction, but they are a
- great tool for anything demonstrative.
- What you're seeing on the right side of the screen here is from a demonstration
- that we did. This was a 3D scan of this patrol vehicle.
- So again, we set the drone down. Um, we set up
- our pillars around the vehicle. We set our ceiling and our
- floor. The drone went off, and it did its thing.
- It came back, and then on board the drone we
- were able to
- create this model. Um, and so this can be helpful for...
- You know, I had a lot of people ask me like, "Why,
- why would we use this mesh model
- of a patrol vehicle?" But as you know, in diagramming and
- oftentimes in anything that's use of force, officer-involved
- type incident, um, they want your diagrams to be pretty
- accurate to the scene.
- And so being able to take these vehicles now, I could do some cleanup and
- get rid of the ground. I can import
- that exact model of a patrol vehicle to position them correctly in a
- diagram to have the correct deliverables in a 3D
- environment to make sure we have the right perspectives.
- This is a really quick fly-through.
- Um, there's not a whole lot to see here except
- for the point cloud that was
- created, but I wanted to give you an idea of what a
- point cloud
- looks like created from the drone images.
- This particular capture was two hundred and twelve images,
- and it took twelve minutes to capture that data.
- Again, that's not twelve minutes of you
- controlling the
- drone. That's just twelve minutes of it going off
- and doing its thing and then
- coming back and landing when it's complete.
- Uh, distance to surface on this was ten feet.
- Our ground sampling distance was .05 inches.
- You can see the overlap and side lap were eighty and seventy percent.
- I took those two hundred and twelve images.
- I imported them into Fero Zone's PhotoPoints
- photogrammetry feature. I set that at medium quality, and
- for the two hundred and twelve images it took about
- thirty-three minutes
- to process on my computer.
- Once it was done, I was able to import that directly back into
- the software and create this quick fly-through video that I'll go ahead and
- play here for you again.
- I will say in this particular instance, I just let the
- photogrammetry engine PhotoPoints, uh, scale the point
- cloud via the GPS coordinates captured by the drone.
- However, we could always make that more accurate
- if we're using ground control
- points or if you have RTK, um, or if you
- have known measurements within multiple images.
- Um, here's a quick image 'cause I get asked
- about this quite a bit, is
- accuracy. Um, this is from a more recent 3D scan, and
- again, I took all of those images, I brought them into FARO Zone, and
- I created a 3D point cloud from them.
- So I thought, you know, what's a quick way to kind of check my
- measurements on this
- scene? So I pulled a wheelbase measurement from,
- um,
- the beloved Ford Crown Victoria police interceptor, which you'll see that
- I got a 9.50 feet for my wheelbase.
- I then went into FARO Zone's, uh, vehicle specs
- database, and I looked up that exact same vehicle, which
- was giving me a
- 9.51.
- Um, and so you'll see that's a difference of
- about one point-- or .12
- inches. Sorry.
- So it's also important to note that, again, this point cloud was just
- generated from GPS coordinates. I did not have a known
- measurement. I did not have ground control points.
- This was just from the 3D scan, importing those images, and doing it
- based on, um,
- the GPS coordinates.
- Using that same data, so the, the big question here is, great,
- I have this drone
- data, like, what do I do with it now?
- Using that same data, I, um, just dropped in some models
- from FARO Zone and very quickly, within minutes, was able to
- generate a 2D diagram of that scene.
- So from the image prior, you saw the police interceptor,
- you saw some evidence
- placards there. I was able to drop those in real quickly,
- added a
- body and another vehicle, couple, uh, identifiers on
- there and some measurements, and literally in less than ten minutes,
- I had
- a 2D diagram that I could potentially attach to a case file, file.
- So the key here is 2D image for your orthomosaic and the 3D
- point cloud for creating
- diagrams as well as creating this perspective because you now have
- elevation.
- So now Russ, I'll throw it off to you so you can talk
- through your case study.
- Do you wanna give some background before we go into the flight path?
- Yeah. And first I wanted to address your, uh, the mesh, the meshed
- model, um, topic. Um, I have, um, I have
- used a meshed models, uh, printed on a 3D printer from scan
- data in the past. Um, and, uh, I think there's no more
- powerful delivery or deliverable for the courtroom is to hand those models
- off to the jury and say, "These are how the, the vehicles came together.
- So, you know, fit that out and, and see for yourself." Um, I didn't
- prepare, uh, uh, some exhibits of those for this, for this,
- uh,
- demonstration, but, uh, I just wanted to let you know
- that that's a very powerful
- tool, and I think it's, uh, it's, uh, more and more commonplace in the courts
- these
- days that they are in fact 3D printing, um, these scans,
- uh, which are converted to 3D, uh, textured meshes and then using them in the
- courtroom.
- Um, the, the, uh, the case study that I'm presenting for you today
- was actually a,
- a double fatality. It happened a few years ago.
- Uh, I'm from New Hampshire, so this is a New Hampshire case.
- Um, this was a, uh, pickup truck, uh, driving, uh, down a hill.
- Um, the operator was, uh, traveling, uh, above the ex, uh,
- posted speed limit, lost control, um, uh, DUI as well,
- um, went off onto the shoulder, came back onto the road, uh, roadway,
- and, uh, into a yaw. And while into, while in a yaw, you know, rotating,
- uh, or yawing across the center line of the
- roadway, uh, struck an
- oncoming, uh, Nissan, uh, vehicle that was going in the opposite
- direction,
- uh, and those occupants, uh, were killed in, in that vehicle.
- Um, the background about the dataset is the police, uh, in this case do,
- uh, do have and, and were using a FARO, um, an older, uh,
- scanner, a FARO Focus S laser scanner, um,
- with a three hundred and fifty meter range,
- um, on that scanner. They took a total of 14
- scans, and this was just about at, um,
- at, um, just as, uh, daylight was, uh, sun was setting on the
- horizon, so things were beginning to get dark.
- They started with, uh, full color and then changed from color
- scan to grayscale scan toward the end.
- Um, I presented-- uh, I've, I've created some deliverables here, one,
- uh, in, in daylight and the other one with basically lights
- out, if you will, or
- lower lighting conditions, just to demonstrate the power of the
- lighting,
- uh, features in the FARO Zone 3D expert
- software.
- Um, so the first, uh, the first set you will see
- is just the police
- data. This was, uh, data given to me by the, uh,
- the investigating
- agency. Nothing more than, um,
- uh, showing that point cloud as created in
- FARO Scene software.
- Um,
- again, as I said, it's 14 scans. You can see that I've cropped the,
- uh, the lower portion of that point cloud.
- Um, the data from, from that point to the, uh, bottom part of your screen
- was very sparse,
- and I didn't, uh, think it was, uh, really worth having as part of the,
- the overall point cloud. It looks a little cloudy on my end, but...
- So this is just a, uh, what we, what we call a fly-through video.
- It was created from the FARO Scene software, where we're looking at the, uh,
- the two vehicles at final rest position in the roadway.
- Of course, this is after the occupants have been
- removed.
- Um, there is quite a, quite a hill there.
- Um, I, I have, uh, I can't recall what the, uh,
- the, uh, elevation was there or the grade on the roadway, but, uh,
- it was fairly
- significant.
- Um, and you'll see when we look at the, uh, some other deliverables here that I've
- created that the, uh, the drone and the laser scanner do a pretty good,
- pretty good
- job of showing the, uh, the, uh, the, the slope of that roadway.
- What I did, um, this was of course a few years ago.
- I had, had gone to the scene again,
- um, as recently as a few months ago with my
- Skydio drone.
- Um, I, uh, I did a, uh, a 2D mapping,
- uh, flight over this scene-
- Um, I believe, um,
- Noreen, you can probably help me with a report.
- Uh, there's, there's the, uh, the point cloud that
- was created from those images.
- I believe it was a total of 56 images,
- and my elevation here was 125 feet above ground
- level at, at takeoff, which is over to the, uh, right where that
- intersection is right there. That was my takeoff point.
- Um, the reason I used that elevation was I had a pretty significant, uh, s- uh,
- grade there with some really, really high trees off on the sides of
- the roadway. Uh, literally, I would say they
- were a couple hundred feet,
- um, high. So I wanted to, uh,
- make sure that I avoided those, um, at least where I was flying and not go
- too far out.
- The purpose of doing this this way was to show that...
- All right, so the police took 14 scans, which basically stopped
- where that double yellow line, um, break is.
- And so I wanted to show the power of adding the drone data to
- the police scan data to expand, um,
- the roadway to show a viewer, um, how much, what the roadway looked like,
- um, in both directions, a little further away from where the point
- of impact was.
- Um, so this was, of course, in a more, uh, a really beautiful time of the year in
- New Hampshire where there was a bit of foliage, but, um,
- so that's the reason there
- are, are colorful leaves out there.
- Um, so this is again from a point cloud
- created in the Ferrozone
- 3D Expert software using the PhotoPoints tool
- and, um,
- 56 photos, 125 feet in elevation.
- Yeah. And I'll just add, add to that, Russ, 'cause I have your report pulled up
- here. Um, those 56 photos only took four minutes of
- flight
- time, so about four minutes to capture what's
- looks like
- 67,000 square feet of space that you captured there.
- I was explaining to the group here that the, the Skydio drone, uh, and you,
- you
- mentioned what it does first when it's doing scanning,
- but it does its sur- I, I
- call it a surveillance flight, if you will, to, to see what,
- what objects are in the area and what it needs to avoid.
- And then when it really does do its thing, it's quite, it's quite fast, uh,
- unnerving fast at, at some, some cases because it,
- it moves around very
- quickly, but, um, does a great job.
- Um, this was, uh, 2D, uh, GPS scanning, so,
- um, not really going down, uh, low to the road or moving around
- objects, just taken from above. And my camera angle for this was
- at, um, 60 degrees on an oblique, so not, not 90
- degrees straight down to the surface, but, um, 60 degrees, uh, oblique
- angle to the surface.
- Yeah. And because I feel like we always get
- these questions too, your overlap was
- 80% and your sidelap was 75%.
- Yes. And my, my go-to is, uh, 75 or better.
- Um,
- obviously the more, the more o-overlap, the, the more data, the better the
- data. But, um, at some point you have to, uh,
- uh, sacrifice data for battery, battery time, um, you know, the
- number of photographs, processing time.
- But, um, this, as you said, was only about a four-minute flight.
- And honestly, I do not recall,
- uh, the processing time on this, but it was not very long.
- 56 photographs in the PhotoPoint software goes, uh, rather quickly.
- Yep.
- All right, so now we can move on to your combined
- point cloud.
- So this is using the, the PhotoPoints tool, um, and then
- combining the, the police data set with the drone data
- set.
- Um, a little hard to see where, where those, where that connection is
- made, and I guess that's a good thing, right?
- So if it's difficult to see where the two emerge together, not a great quality
- video, but, um, where those are merged together is very
- difficult for you to see, meaning that the data was really, really tight.
- The drone data was, um, the accuracy of that data was really
- great, so there was no difficulty in aligning w-
- that with the scan data
- captured by the Focus scanner.
- Um, the break in the asphalt there going from
- dark to light, by the way, is not
- from the two data sets. That was new asphalt versus old asphalt,
- so that's, uh, the cause of that. Um,
- so what it, what this allows you to do is go out to a scene at a later time,
- let's say with a drone, and then,
- um,
- fly over areas that you wish to add as part of your, your
- overall view or your presentation in the courtroom.
- And so you can take what, what would normally be, or in this case,
- a smaller data
- set by police with a laser scanner and add to that the drone,
- uh, data and make for a larger, more, more complete presentation.
- Yeah, I think, um, one of the
- biggest questions that I've had over the last couple years
- is, if I have a laser
- scanner, why would I also be using drone?
- And I think the idea here is because the drone capture is so fast,
- you can capture the whole roadway, right, with a drone, and then you could
- just focus your laser scanner on the vehicles or the important evidence.
- Um, and same goes for, you know, I've worked plenty of shootings that have
- gone, like, blocks, right? A drone can capture that data really
- quickly, and then we can just focus on locations
- for potential
- trajectory with laser scanners. So it's a time saver.
- Um, you're not sacrificing a whole lot of
- accuracy there.
- You're getting o- out of the scene faster.
- Um, you're gonna be able to process this back.
- And then you can always merge it, and it doesn't always have to be merging
- it with
- laser scanners, right? With Ferrozone software,
- because it's data agnostic, um, you can bring in total station
- data and align that with your drone image.
- You could bring in hand measurements and align that with a drone image.
- I mean, preferably you'd be moving away from hand
- measurements, and you'd be using
- other technologies. But you can always just supplement
- whatever
- you're currently doing with the additional images or thermosaic or
- 3D point clouds from the drone as well.
- So, um, so the best part to me about Ferrozone is,
- yeah,
- it's called Ferrozone, but it's data agnostic.
- You can really just bring anything into it and create your j- courtroom
- deliverables.
- True. And the other thing that, that I find,
- uh, one of the, one of the, the benefits that I most like is the fact
- that I can get information from above that my scanner cannot
- see.
- So the, the laser scanner is, um, more or less a
- terrestrial, uh, instrument that captures data from the ground,
- does a,
- a fantastic job with that, but, um, there are just times when
- it's not, it's not practical to get that laser scanner in a place that's
- high enough to see things from above,
- and that's where the drone really shines.
- Um, so I, I, I preach the, you know, it's the benefit of having both
- tools in the toolbox, one for capturing things from high above,
- um, filling in, say, rooftops and areas that the dr- that the
- scanner cannot see. And vice versa, you mentioned it earlier on in the
- presentation that the use of, of scanners, drones inside is
- not always practical. Um, then that's where the laser scanner shines
- as far as doing, uh, very, very, um,
- high resolution measurements, uh, great accuracy inside smaller
- spaces.
- This is, uh, an overview, uh, actually not an ...
- Well, it's an overview too, but this is an orthomosaic,
- um, diagram created from the, the, uh, combination
- of both the police scan data and the drone data.
- Um, so both point clouds here are visible and merged into one, and
- this of course was done in the FARO Zone, uh, 3D, uh,
- expert software. Um,
- and then, uh, from that, um, you showed earlier on, um, I was able
- to create a line drawing. Now, this line drawing's not as fancy as most
- line
- drawings that I do, but I, I just simply wanted to
- demonstrate how easy it is to just take that point cloud
- information,
- those 3D data points, and then using the FARO Zone
- software, connecting all of those data points
- and creating a scale,
- an accurate and scale diagram from, from the combination of the
- drone and scan data.
- And then I, I took it a- another step further here and, and as I
- indicated, uh, you know, introduced a nighttime scene where I have
- my,
- my Nissan coming toward me, the truck losing control and ha-
- having a head-on
- collision and then rotating out to final rest.
- And I, I didn't remove the vehicles at final rest
- on
- purpose just so you could see these cars when they're, when they're at final rest,
- they're actually moving down to the places where they were actually found.
- Another step, had I taken it to the next level in FARO
- Zone software, I could
- have, um, damaged those vehicles to fit the scan
- damage that was captured by the police.
- Uh, in this case it was just for the purpose of showing you
- that not only can we do
- the merging of point cloud data, but we can also animate
- this collision, uh, and have these vehicles, uh, uh,
- demonstrate how these vehicles hit one another and then rotated to final rest.
- We can have them crushed, and in this case we can even make it a
- nighttime scene
- where we've,
- um,
- made it darker, made the point cloud darker,
- and then added the use of headlights
- and taillights, or in this case, uh, no brake lights, but, uh,
- brake lights are,
- uh, another thing that we could have done with this,
- or streetlights even if there
- were some there.
- So that's one angle.
- Can I-
- Yes, go ahead. Sorry, Noreen.
- Oh, no, sorry. I just wanted to know if you could speak to the heads-up display
- up
- in the corner for those who aren't familiar with creating these animations in
- Zone.
- Oh, yeah. Thank you for that. Um, yeah, I, I, I like the heads-up display.
- Um, in, when you're creating an animation in the
- software, you can
- identify where the point of impact is.
- And then so of course point of impact is, in the crash reconstruction world, time
- zero, and everything goes back from the point of
- impact, uh,
- in distance and in time. So what, what this is
- showing is this, by the way, is set up just to, uh,
- demonstrate where the
- Nissan was. I could have turn up- turned on a heads-up display for both
- vehicles,
- but, um, this case I wanted to show how far out
- was the, uh,
- the Nissan, how lo- how long was it to the time of impact, and the
- distance traveled to the point of impact, and the velocity of the vehicle at the
- time of impact. In this case, um, the velocity of the Nissan was
- 35 miles per hour. The truck was about 50 miles per hour,
- um, give or take. Um, that was based on a calculation that was also
- done in the software using, uh, uh, a head-on or
- inline, uh, momentum calculation, but that was,
- of course, that's not part of this demonstration.
- But that was done in the FARO Zone software as well.
- Yeah, and I think you ... I, I mentioned this earlier,
- but I think the key point
- here is that
- Zone has made it so easy to just do everything in Zone.
- Um, one thing that was super frustrating to me being out in
- the field was that
- everything seemed to take so many pieces of software, and you had to be relatively
- proficient in all of them. Um,
- and sometimes that could be overwhelming, especially if you weren't using some of
- them as frequently as others. And so Zone has made it really
- easy where once you bring in your data into Zone, whatever that is, right?
- That might be your laser scanner data, that might be just drone images,
- that might
- be drone point clouds, doesn't matter.
- Um, from there you have all the tools that you need to go from
- that data to a simple line drawing, to an advanced
- reconstruction, to an animation, to a fly-through video, and so on
- and so on and so
- on. Um, and I think it's just excellent to have
- all your tools in one
- place so you don't have to find all these other pieces of software to kind of get
- you all the answers that you need for your case.
- Yeah, and there's another, another wonderful tool,
- and of course it's beyond the
- scope of this, this, uh, workshop, but, um, the new, the new feature called
- Mass Zone, which is, um, our, our, uh, computer,
- uh, prediction simulation, um, component of the software that allows
- us to not only, you know, animate how those vehicles come together,
- but to,
- but to, with, you know, changing the parameters, see how
- that crash would unfold,
- um, uh, in terms of its velocity, final rest positions,
- rotation, and so forth. So a re- really powerful new feature in the
- 2024 software. So if you have a, a moment, I would encourage you to,
- to, uh, take a look at that.
- Excellent.
- Um, I think I wanna talk about real quickly,
- um, from a Skydio
- perspective, we will be doing a live demonstration of
- 3D
- scan. I know Russ touched on this briefly, um, but what makes
- 3D scan great and what makes, um, Skydio easy to use is
- our advanced AI and autonomy. It's really, really hard, if
- not difficult, to just crash these drones, right?
- They're constantly building the environment around themselves.
- Um, and I've even seen Russ, I don't know if it was last year
- or the year before,
- just showing people, like, "Watch, I will try to, to fly this right into this
- wall and it will stop." Because it's building the environment around itself.
- It, it knows where it is in space and it knows where, for example,
- your vehicles
- are, are, are in space. So when you're
- setting up your volume, your pillars to scan whatever it is that
- you need in 3D, it's not gonna hit anything 'cause it sees
- them, and
- it's building this environment internally for itself so that it
- does it automatically, autonomously, without you having to put much
- effort into it other than telling it what to do.
- So if you would like to see a demonstration of that,
- we have a warehouse,
- um, out in California where we have set up
- some crashed vehicles.
- One of our solutions engineers will be live there with our
- drone to show the process of the 3D scan.
- So previously I showed you kind of a quick sped up video of what it looks
- like to set up your pillars, set up your volume, choose some other, um,
- settings
- in there, and then have the go- the drone go do its thing,
- but you'll be able to
- get to see that live. In addition to that, we will provide you with
- all of the images captured by the drone.
- So you can go ahead and bring them into Zone
- or whatever software you're using
- to see what that point cloud looks like, to see how long it takes to process, and
- to see what you can do with it once you've created that point cloud.
- Um, so the QR code here is, uh, for registration for that.
- That will be on February 2nd from 1:00 to 2:00 PM Eastern.
- Um, and uh, we will also send out an email.
- So if you see an email pen, uh, post this webinar, it is for that
- live demonstration. But come hang out, see what it's all about.
- Um, you know, we've had some, a lot of interest in flying
- directly in our, our warehouse and beyond crash reconstruction.
- If you're interested in how, um, remote operations work, you can
- always, um, have a demonstration where we give you a
- link and you can
- fly the drone from wherever you are in the world, although the drone
- is located in
- a warehouse in California. So, um, a couple of exciting things
- there to do. Um, and next we're gonna go to some
- Q&A. Um, I've also included both of our email
- addresses here in case you need to
- reach out to us for anything.
- Um,
- one question, Russ. We'll just go to this one quickly 'cause
- it's the last thing
- that you talked about, but how much time do you have in creating
- those
- deliverables?
- Uh, not that long, really. The, the, the longest time is for the processing of
- the data, uh, to create the point cloud from the images,
- and that,
- that depends largely on the number of photographs that are taken.
- Um, I would say on a, on a, let's say...
- I'm, I'm throwing some numbers out and really I'm, I'm estimating.
- Let's say 100, 100-photograph project, so a project where the drone
- has captured 100 photographs. Processing time, I think you would
- sugge- you had said this about one earlier,
- maybe close to 30 minutes, uh,
- perhaps less. There are choices that you can make in the software on
- whether you want to choose a low resolution, medium, or high
- resolution point cloud. Um, so, so of course the higher the
- resolution you choose, um, the longer that process would take.
- Um, but the process of merging the two
- datasets, um,
- is literally, uh, long enough for you to click three points
- in
- one dataset
- and similar or common three points in the second dataset,
- and then they would, they would then align.
- And lastly, you would simply just, uh, um, create that merged, uh,
- point cloud, which can again take upwards of,
- uh, I, I'm gonna say 10 or 15 minutes for the same 100-photo
- dataset. And again, that's an estimate on the numbers,
- so please don't quote me on
- that, but somewhere in that area.
- Excellent. Um, yeah, and I think just in all the other
- deliverables, it's gonna
- depend on
- how proficient you are with the software.
- Russ can do things a lot faster [laughs] than I can because he's just
- more proficient with it. [laughs] Um, but I can do a shooting reconstruction in
- minutes, so there. Um, but I think it just depends on your
- proficiency in the
- software and, you know, how familiar you are, not only with the
- tools that you're using, but the concepts and the science
- behind what you're trying to produce in there as well, so kind of-
- And I, I'm looking forward to follow up on that.
- I'm looking forward to getting data from the new X10 because I, I know
- that the sensor on that, that, um, drone is like
- crazy, over the top fantastic. And I can imagine, and the reason
- I'm going there with this, is that, that the data is more-- there's much
- more saturated data from those point clouds than there would be from,
- say, a, a drone that doesn't have a sensor quite
- as capable as that.
- And so there would be, of course, more data to process,
- and I'd be curious to see
- what that looks like. But, um, as far as I'm concerned, more data, um,
- especially s- uh, 3D scan data where it's filling in the
- sides of objects is, is a,
- is a good thing. Um, and not, not having sparse data.
- Sparse data doesn't look so great.
- Yep. Agreed. Um, in creating point clouds, where is the data
- stored, cloud or local? [clears throat]
- Great. For us, it's all local. Um, for Fero-
- Ferozone does not-- we currently, for public safety anyway, we do not,
- we do not store data, uh, in the cloud.
- Um,
- m- all of your, all of your processing, all of your workflow, everything occurs on
- your local, uh, either your laptop computer
- or your desktop.
- Or if, if you have a networked, uh, uh, system, then that would work as well.
- But none of it goes to the cloud, it's all locally based.
- So Skydio is whichever you choose.
- Um, we have Skydio Cloud where you can send
- all your data there.
- We also have a partnership with Axon, so you can send your data to
- evidence.com. Um, it just goes to Skydio Cloud
- and then straight over to your
- evidence.com account. Or you can just save it locally to the SD
- card, pop it out, and do with it whatever you need to do,
- put it in your own
- systems, um, in-house or on-prem.
- Um, how are the deliverables transferred
- to prosecutors?
- Is that for you or me? [laughs]
- Either or. I mean, I think, I think that there's a couple ways you can go
- about
- this, right? Um, if you're just talking about drone images,
- you're gonna
- put those into evidence just as you would your scene photography.
- So those are gonna get to your prosecutors in the same way.
- Um, hopefully not on a DVD drive anymore.
- Hopefully [laughs] we're a bit past that.
- Um, but if you're doing deliverables in the
- software, specifically
- FARO Zone, and I'll let Russ speak to this more,
- but you could use a Zone ToGo.
- Yeah, so now there are a number of different ways.
- Um,
- so depending on the size of the final product, right?
- So, um, size is certainly going to dictate what,
- what sort of, uh,
- device you would need to hold the capacity or the size of that, that
- file. Zone ToGo, which is the, the
- FARO Zone version of our Scene ToGo, um, uh,
- component of the Scene software is, is very, very good.
- It doesn't, it does not require the, uh, the user or the
- viewer to have separate software.
- It's all, it's all run, uh, from a thumb drive,
- um, or an SD card or any other device that you're handing off, an
- external drive, let's say.
- Um, so it, it can be exported just like any other
- file really, uh,
- as long as that device is capable of, of holding, uh, the large dataset.
- In some cases they can be very large.
- But, um, storage is pretty inexpensive these days
- compared to what it
- used to be.
- Um, a one terabyte, uh, SSD, um,
- can be purchased for under $100, um, two terabyte for
- under $200. So, uh, storage is no longer really much of an issue.
- Um, it is difficult, I will admit, to try to transfer this information,
- uh, in the cloud, uh, online because of the file sizes.
- Um, that's where the Zone ToGo comes in.
- The file sizes are much more, um, reasonable or manageable than,
- than, say, handing over the original dataset.
- So original files versus copies of files, there can be a difference in, in that, in
- the sizes of the files themselves.
- It would be exported-
- Yeah
- ... it, it would be exported in a, a various number of formats.
- If it's just the point cloud, it can be exported in FARO Zone
- as an E57, a PTS, an LAS, um,
- or in FARO Zone's, uh,
- um,
- uh, format a .fzproj file. Um, so,
- and those are fairly common ways of handing them over to other FARO customers.
- Yeah, excellent. I think, um, if, if Zone ToGo is a phrase
- that's, uh, foreign to you and you are a Zone user,
- you need to learn all about it.
- Um, I heard recently a couple agencies say,
- "We have all this great data.
- We have no way to get it to our prosecutors." And you,
- you absolutely can do that,
- um, with Zone ToGo, and you can package it all up in a nice
- little thumb drive for
- them and pass it on, and they can use it without needing
- software and
- without needing licensing. So, um, make sure you familiarize yourself with
- that if you are not aware.
- Um, the next question here, and I hopefully am interpreting this
- right, um, will it capture data at night?
- Um, I will speak to this briefly, and then Russ,
- if you've got something to add on
- there.
- Um, so currently the X10 specifically
- is, um,
- at first customership, we were shipping it with one sensor package.
- However, the X10 has what we're calling an
- upgradable sensor
- package. So we will create more of those during the
- lifetime of the X10.
- And shortly to come out in the new year is a new sensor
- that is going to be best for mapping,
- and it also includes a, um, an, an
- embedded LED flashlight on the sensor.
- And so we're doing some testing right now to see if, because X10
- or because Skydio drones can fly so close to the subject
- without crashing, is that light enough to
- provide the light that you need to capture this data at
- night?
- So that's kind of still under testing. I'm not gonna say yes or no yet.
- However, a lot of our agencies that are using drones are just bringing
- out some lights to illuminate the area that they need, and then they're still
- flying their drones for data capture.
- So you can, you just have to be a little bit more
- creative because you have to
- remember it is a camera. Um, and so likewise with
- any other photography, low light conditions are just difficult.
- And you beat me to it, Noreen. I, I, my simplest answer for that is
- it is a camera. At the end of the day, that's, you know...
- So ask yourself that question. If you, if you took
- a photograph from the height of where the drone happens to be, will
- that image, will, will you be able to see things?
- Is there enough available light or do you need to introduce artificial
- light
- into that scene to capture that image?
- I have scanned at nighttime, uh, in the past
- in an area where there's a lot of streetlight or a lot of ambient light
- from, uh, businesses, say along the side of a busy road.
- Uh, that is doable, but of course also with cameras you have
- to take
- speed into consideration. The shutter speed would need to
- be, uh, I mean, it would need to be open longer,
- therefore your movement would
- have to be almost no movement at all, or if there is any movement, very,
- very slow movement. So in nighttime photography cases you
- would
- have to,
- um, have your drone stop, take a picture, move, stop, take a picture,
- move, and so on in order to capture really great
- images with low light.
- Yeah, and that actually kind of ties into
- another question that was on here about,
- um, does the shutter affect motion blur?
- And, um, you can always, and, and 3D scan has this ability too, to
- stop at structure. So it will intentionally stop to capture
- that image before it
- moves on. Um, it's like, you know, a little bit longer,
- but you're gonna get the
- best data possible.
- Mm-hmm.
- Um, a question here for you, Russ. What aerial perspectives do you think work
- best for crash data?
- Um, if I'm understanding the question
- correctly,
- um, I, I typically, my, my go-to perspective, I
- go to-- I, I try to, of course, always be above
- the highest object in my, in my area that I'm trying to fly with-
- in. And, um, and generally speaking, um, I'm even in New
- England where I have a lot of mountains and trees, I-- if I, I try to take the
- highest, uh, point for launch since that's the
- measurement from my
- above ground level where my, my drone will go.
- I generally can go up to about a hundred and twenty-five to a hundred and fifty
- feet above the ground.
- Um, my camera setting is, um, as I said earlier,
- FARO Zone re- or maybe I didn't say it earlier in this
- workshop.
- Maybe it was in a class I was teaching.
- But, but the, uh, the FARO Zone software requires,
- um, an oblique angle, uh, setting for the camera.
- Uh, it cannot, uh, do a, um,
- a neutral directional. Uh, in other words, it cannot use a
- ninety-degree, uh, uh, images to the surface.
- Those do not work well with the software.
- So that camera angle I recommend is a sixty-degree oblique.
- Um, and really anything greater than zero will
- work, but sixty
- degrees seems to be a sweet spot, at least, uh, from my experience.
- Um, I set my overlapping, uh, data, uh, my overlapping
- images at either seventy-five or eighty percent, um, as, as
- a,
- a ba- a baseline, if you will, and sometimes even more than that.
- But seventy-five to eighty is, is typically my go-to.
- Hopefully that answers that question.
- Excellent. Yeah. And for reference in the, uh, drone data that you showed in that
- case study, uh, your, your gimbal angle
- was sixty degrees for that.
- Yeah, I usually use a cross grid pattern.
- I don't believe that I flew a-- I might have thrown a, flew a cross
- grid pattern there.
- Um, basically I send the drone down on a
- lawnmower pattern
- back and forth, and the cross grid is then reversing or, or
- rotating that, that flight ninety degrees to the original
- path, and
- then doing another series of flight, another flight, if you will, over the same
- area, but from a ninety-degree angle.
- Um, and that gives me the best data, especially with the oblique.
- The drone is capturing data, looking at all areas of the objects
- below, uh, from different angles as opposed to just
- down and back.
- Uh, so I find that that produces a much nicer point cloud.
- Excellent.
- Um, the question here is, should one mark pillars near open doors of
- a vehicle for
- best results? Um,
- so basically for, for 3D scan, you can put your pillars
- anywhere so long as everything you need is inside the pillars.
- Um, if you ha-- I mean, we have indoor capture for
- 3D scans. You could set an entire, you know, pl-- uh, building or a
- warehouse for capture. Um,
- as long as your pillars just encompass everything you absolutely need,
- you will have the results that you need for the 3D scan and the, um,
- subsequent point cloud.
- Um, Russ, how often are you flying manually?
- Oh, we were just talking about that today.
- Yeah. [laughs]
- Um, my go, my go-to is autonomous. I usually do a-- I have a, I
- use a mapping application, uh, when I'm using the Skydio.
- Of course, I'm using the enterprise application,
- and that pretty much, uh, handles
- everything for me. I tell it how high I want it to fly
- and what my camera angle's
- going to be and, and set my, my, uh,
- my parameter on, on the, uh, the map that's provided,
- and I hit go.
- Um, but there are times when I'm looking for information along
- the sides of objects such as vehicles at a, at a collision where I wanna
- fly around that,
- or I wanna capture the sides of buildings, and I wanna be able to get the facades
- of buildings, uh, more clearly and get more data points for that area.
- Maybe I wanna model it. Um, in that case, I would, I would typically do
- a separate flight where I'm flying low and in, in-- With the, with the
- Skydio drone, it's easy because I can do a 3D scan from
- that, and as you said,
- set the parameters of what that scan needs to be.
- And I know that the drone's gonna take whatever number of images it needs
- to capture that object. Um, larger, of course, takes more
- time, more photos. Um, if I wasn't using a Skydio drone
- or if I was flying manually even with the Skydio drone, I would drop the height,
- of course, my altitude down and take a series of photographs along
- those, those faces of objects that I wanted to be, be sure I had,
- uh, data for.
- So it just depends, but, um, I'd say percentage-wise, not very often.
- Um, with the s- with the oblique on the camera,
- I'm capturing
- pretty much what I need, at least at a crash scene.
- Um,
- if I wa- if I wanted to do a model of a car
- and then 3D print that car, it
- would probably-- I'd probably take, uh, I'd do a 3D scan and have
- that, the Skydio drone at a much lower and closer, uh, observation
- point, uh, than, than high above.
- Excellent. And then just to add to that, with 3D scan, um,
- once it goes around and it does its thing and it comes back,
- you can view that on
- the controller. And if you see that you would like some more data in
- a certain area, you can go back around and take additional manual photographs,
- and those will just all go into the same flight project, essentially.
- So you'll just have additional photos in the areas that were more important to you.
- Let's say,
- let's say you set up, um, your 3D scan around two
- vehicles, but you wanted more data in an area of
- crush damage.
- You could manually take those images after the 3D scan flight just to
- supplement what it had captured.
- I just wanna put in a plug for the, for the meshing, uh, component of your,
- of your drone. It's something that you don't see.
- Um, in fact, I don't know of any other, uh, manufacturer that's doing
- that. Uh, maybe they are, but I'm not aware of that.
- And that's a whole another process beyond the creation of the point cloud
- And beyond that, the, the combination of the point clouds to,
- to create a me-- a
- textured mesh of an object, um, I use that, uh, pr-
- procedure all the time. Um, so to see that your software does that or y-
- your, uh, your solution does that is, is a good thing for me.
- Yeah. We're really excited about the onboard
- model, so excited to see what people
- do with it.
- Um, we have two more questions on here.
- We have a few more minutes, so if you have anything else, go ahead
- and throw it in
- the chat there. Um, but Russ, can you just quickly talk about FARO
- handheld scanners? Um, I'm sure they're talking about the
- Freestyle 2, and
- I don't know specifically for what, but maybe we'll just talk about it in the
- situation of crash.
- By the way, um,
- Noreen is a Freestyle 2 expert, [laughs] or was.
- [laughs] Um, but you now work for Skydio.
- But, um, so the Freestyle 2 hand scanner, an amazing device.
- It's, um, come a long way, uh, since I started with FARO
- seven years ago, almost eight years ago.
- Um, it is a portable device that, that can be, uh,
- that can be moved around objects. Um, uh, large objects or
- larger objects such as vehicles, um, in small rooms can be
- scanned in a matter of a minute or two or three.
- Whereas if those objects were scanned using the, the, uh, laser
- scanner, that might, might take a bit more time than that.
- Um, at the end of that, uh, the capture, um, it's using a process of
- LIDAR photogrammetry to create a three-dimensional point cloud.
- That data can then be, um,
- uh, merged easily with, um, scan data captured by the, uh,
- FARO Focus scanners, or our Premium scanner.
- So, um, creates some really nice, uh, colored, uh, point clouds
- for, um, bodies, let's say ... Again, I wanna emphasize
- smaller spaces,
- um, uh, small rooms, small offices, uh, around cars,
- around bodies, and, and that can then be easily merged in with,
- with, uh, scan data captured by our Premium scanner.
- Um, not something that works, uh, well, um, in
- direct sunlight, um, so it's, it does have a limitation where it
- can't be used in direct sunlight outside, um, in bright, on bright,
- sunny days. It does a fairly good job in, in, uh, scanning
- objects that are covered or where, where direct sunlight is blocked from,
- from reaching the object.
- The accuracy-
- Excellent
- ... is as, as good as the laser scanner is.
- Yeah. Hopefully that answered the question
- there.
- Um, and again, that's another thing that,
- you know, you could even use a drone for the
- majority of your scene and then the
- Focus for the ve- or, I'm sorry, the Freestyle 2 for the vehicles.
- It's the same process to merge those point clouds in Zone
- as it would be for the Focus.
- [clears throat] Um,
- okay. Next question. Um, and this is specific
- to Zone software. They said their agency has Zone 3D
- Advanced, and they were told that would work for importing drone
- photographs, but they're having issues.
- Um,
- just a quick call-out on here. You need FARO Zone 3D
- Expert.
- Mm-hmm.
- So it's, um, likely that you just don't have the most updated
- version of the software. It was the ...
- Let me not get my years messed up here.
- Two years ago, I think.
- Two years roughly. Is that sound right?
- La- last year.
- Last year.
- So the 2023 version, um, about this time last year, uh,
- of FARO Zone 3D Advanced became 3D Expert,
- and that's what allowed for PhotoPoints, which is the photogrammetry feature.
- Mm-hmm.
- So you're just gonna have to confirm if you have, um, the ability to get
- Expert, and then you will have, um, the ability
- to use PhotoPoints.
- And if you do have the, the, a version that does do that,
- what I'm seeing
- at least a couple of times recently where customers believe that you
- simply just import the, the images, and that creates a point cloud.
- Well, there's a process that, that needs to be worked through.
- There's a workflow that needs to, uh, get you from images to a point
- cloud. And, um, it's not difficult, but it's just a process, and you need to
- know what that process is.
- Excellent.
- Um, and then it looks like we have one more
- question, and then we can wrap up.
- So the question is how large can the area be for the 3D scan function?
- And I don't know that I have a specific number for
- you. Um, however, with 3D scan, we created
- indoor capture, which was intended to be used in
- warehouses or factories, like really large indoor
- environments.
- Um, and really what you're restricted by there
- is just the battery flight
- time, like how, how often you're just gonna have to switch
- the batteries throughout
- that process. Um, but I don't know that I have a specific like it
- won't capture beyond this, um, 'cause I don't know that anyone's said
- yet, "Hey, I tried to scan this, this large of an area
- and it just didn't work."
- Um, again,
- larger areas, more images, lots more data,
- um, so you're talking about really, really big data sets at the end of that.
- [clears throat] All right. Russ, well,
- we've come to the end of our session.
- Thank you so much for hanging out with me today. I really appreciate it.
- I've missed hanging out with you, especially in webinars.
- Um-
- Yeah, but everyone asks for you all the time,
- Noreen, so you have to, we have to do
- this more.
- [laughs] Yeah, we will. Um, to everyone that's hanging out with us
- today, really appreciate it. Hopefully you kind of took away some
- nuggets of
- knowledge along the way, whether it's about drone use
- or the use of s- uh,
- FARO Zone software to use all of this technology to create your
- courtroom deliverables. Um, we just put up ...
- On the Skydio end, we just put up a crash
- and crime scene documentation page on our
- website where you can get some more information there
- and hopefully some more cool
- projects to follow as we start to do some more work here with our new X10.
- If you have any questions, you can reach out to us.
- You know where to find us.
- Otherwise, thank you. We really appreciate you hanging out and-
Watch Time
54 Mins 53 Secs
Related Webinars
Pre-Incident Planning for FIFA: Turning Drone Data into 3D Operational Readiness
Watch webinar
Drone operations for FIFA host cities: Planning for zero-failure security
Watch webinar
Drone as First Responder: Changes in Policing, Changes in Prosecution
Watch webinar