RTC360 Inclinometer/IMU accuracy
-
- I have made <0 posts
- Posts: 2
- Joined: Wed Mar 03, 2010 3:00 pm
- 14
- Full Name: Spike Milligan
- Company Details: Mimic
- Company Position Title: President
- Country: USA
- Linkedin Profile: Yes
RTC360 Inclinometer/IMU accuracy
Hi there,
I know some threads covered this briefly just after RTC360 launch. I was wondering if anyone verified tilt sensor/imu level accuracy on the RTC360. We're seeing better than 3' published in a registered data set of 300 scans of a large open building interior. We're not seeing any info exposed in Register 360 or Cyclone to be able to see the tilt sensor numbers or transformations. We will analyze the data more and come up with a more exact number but it appears better than 1'.
Has anyone analyzed it further?
Thanks
Spike
I know some threads covered this briefly just after RTC360 launch. I was wondering if anyone verified tilt sensor/imu level accuracy on the RTC360. We're seeing better than 3' published in a registered data set of 300 scans of a large open building interior. We're not seeing any info exposed in Register 360 or Cyclone to be able to see the tilt sensor numbers or transformations. We will analyze the data more and come up with a more exact number but it appears better than 1'.
Has anyone analyzed it further?
Thanks
Spike
- gsisman
- V.I.P Member
- Posts: 898
- Joined: Fri Oct 07, 2016 1:51 pm
- 7
- Full Name: Steve Long
- Company Details: Montgomery County DOT _ MD
- Company Position Title: Land Survey Supervisor
- Country: United States
- Skype Name: gsisman1
- Linkedin Profile: Yes
- Has thanked: 767 times
- Been thanked: 149 times
Re: RTC360 Inclinometer/IMU accuracy
Both our crew and another consultant using an RTC360 to scan a parking garage have noticed the tilt issue. He brought up a good point. In both the Pegagsus and the backpack mobile scanners you have to run a pattern of sorts before the IMU really kicks in. He felt that scans that came out tilted had been ones that were turned on and scanned right a way.Like switching to another floor or after a break in starting a new bunch of scans. I had some that definitely lost the sense of "up" when I went to long with the VIS. I didn't notice if ones from a "Cold" start were tended to be tilted, but he felt that the longer the scanner wa being moved around the better the sense of tilt correct itself. Has anyone else took enough notice of WHEN their scans got tilted with that instrument. Just wondering whether it's a procedural thing that we are not aware of.spike3d wrote: ↑Mon May 13, 2019 9:51 pm Hi there,
I know some threads covered this briefly just after RTC360 launch. I was wondering if anyone verified tilt sensor/imu level accuracy on the RTC360. We're seeing better than 3' published in a registered data set of 300 scans of a large open building interior. We're not seeing any info exposed in Register 360 or Cyclone to be able to see the tilt sensor numbers or transformations. We will analyze the data more and come up with a more exact number but it appears better than 1'.
Has anyone analyzed it further?
Thanks
Spike
-
- V.I.P Member
- Posts: 643
- Joined: Mon Aug 11, 2008 4:13 pm
- 15
- Full Name: James Rye
- Company Details: Merrett Survey Partnership
- Company Position Title: Spatial Analyst
- Has thanked: 28 times
- Been thanked: 69 times
Re: RTC360 Inclinometer/IMU accuracy
We use the RTC with targets because so many scans have come in "wonky". Have discussed the issue with Leica and SCCS who say Cyclone isn't reading the tilt sensor properly in some cases. We've been advised that a new version of Cyclone is on the way to resolve the issue. Currently, we are advised to press the button that looks like an 'R' in the visual alignment window to switch between levelled and unlevelled (or revert to the 3 point picks).
When doing a large building survey, I'd always do the stairwells with a P series scanner and/or shoot in control points with a total station.
If it is, then we are making the same procedural error.
When doing a large building survey, I'd always do the stairwells with a P series scanner and/or shoot in control points with a total station.
If it is, then we are making the same procedural error.
- pburrows145
- V.I.P Member
- Posts: 2152
- Joined: Wed Sep 12, 2007 9:16 am
- 16
- Full Name: Paul Burrows
- Company Details: Leica Geosystems Europe
- Company Position Title: Scanning Solutions Manager - Europe
- Country: UK
- Linkedin Profile: Yes
- Location: UK
- Has thanked: 109 times
- Been thanked: 123 times
Re: RTC360 Inclinometer/IMU accuracy
FW and SW improvements on the way to resolve the issue in a tiny fraction of cases. Also, a new tilt function has been added to the next release of Cyclone REGISTER 360 (as in FIELD 360) to quickly resolve this.Jamesrye wrote: ↑Tue May 14, 2019 8:43 am We use the RTC with targets because so many scans have come in "wonky". Have discussed the issue with Leica and SCCS who say Cyclone isn't reading the tilt sensor properly in some cases. We've been advised that a new version of Cyclone is on the way to resolve the issue. Currently, we are advised to press the button that looks like an 'R' in the visual alignment window to switch between levelled and unlevelled (or revert to the 3 point picks).
When doing a large building survey, I'd always do the stairwells with a P series scanner and/or shoot in control points with a total station.
If it is, then we are making the same procedural error.
-
- V.I.P Member
- Posts: 643
- Joined: Mon Aug 11, 2008 4:13 pm
- 15
- Full Name: James Rye
- Company Details: Merrett Survey Partnership
- Company Position Title: Spatial Analyst
- Has thanked: 28 times
- Been thanked: 69 times
Re: RTC360 Inclinometer/IMU accuracy
Thanks Paul, that's good to know. For us, it isn't a tiny fraction of cases, it's every job. And for every 100 scans, we are getting 5-10 that are "wonky".
-
- I have made <0 posts
- Posts: 2
- Joined: Wed Mar 03, 2010 3:00 pm
- 14
- Full Name: Spike Milligan
- Company Details: Mimic
- Company Position Title: President
- Country: USA
- Linkedin Profile: Yes
Re: RTC360 Inclinometer/IMU accuracy
Thanks for the feedback. Looking forward to the next release of Register 360. On a side note any thoughts on a procedure for optimizing level in a scan data set when not using a P series or survey control points for level. With no ability to lock a specific scan in Register 360 I'm not sure that using an initially tripod leveled scan can work.
-
- I have made 50-60 posts
- Posts: 52
- Joined: Thu Jun 06, 2013 2:50 pm
- 10
- Full Name: Shaun Lewis
- Company Details: Clark Construciton
- Company Position Title: Senior Manager - Reality Capture
- Country: USA
- Linkedin Profile: Yes
- Been thanked: 1 time
Re: RTC360 Inclinometer/IMU accuracy
For interior work utilize a 360 laser level and set 3 or 4 targets at the same elevation as a check. If it does come in wonky you can use those targets in a new registration to level it out.Thanks for the feedback. Looking forward to the next release of Register 360. On a side note any thoughts on a procedure for optimizing level in a scan data set when not using a P series or survey control points for level. With no ability to lock a specific scan in Register 360 I'm not sure that using an initially tripod leveled scan can work.
-
- I have made <0 posts
- Posts: 6
- Joined: Tue May 14, 2019 6:56 pm
- 4
- Full Name: Tim Bennett
- Company Details: Green Seal Environmental Inc
- Company Position Title: Director of Land Surveying
- Country: United States
- Linkedin Profile: Yes
- Has thanked: 1 time
- Been thanked: 2 times
Re: RTC360 Inclinometer/IMU accuracy
Is there a way to set the elevation of a target without having an x and Y coordinate for it? I'm kinda new to laser scanning, but have been surveying for 30+ years. I've done a few scans with our new RTC360 and my outdoor scans with 3D survey control on our targets work out fine. I import the XYZ control from a csv text file and then register my fixed scan to it, no problem.
But now I have a new project and it is in a 25 ft x 15 ft room, deep in the basement of a hospital, where among other things I need to provide a floor flatness and floor levelness report. This had to be a super tight survey to determine the intersection of the 3 axes of medical imaging machine within 3 mm, so I set 14 targets on the walls and 4 flat on the floor and then did 5 scans, one in the middle and one near each corner of the room ( I probably could have done it in 2 or 3 scans or possibly even one, but at 4 minutes per scan and a surveyors compulsion for redundancy, might as well get at least twice what I needed...) I then used a digital autolevel to run through the four targets on the floor to get true elevations on them to within about 2mm. I did this because knowing that the RTC360 does not have a DAC I wanted to have true level data in the end.
So I registered all my scans together and then exported the control vertices to a text file, and edited the text file and changed the Z coordinates of the targets to the leveled elevations. Then I imported the edited text file into Cylclone and registered the unified scan to the new leveled control. It worked out pretty good and I found that the "unleveled" scans had a tilt of almost 0.02 ft across the 25 ft length of the room, which I was able to correct with this method.
I am wondering if there is a better way to do this - I feel like my method was a bit bass ackwards in that I had to basically create fake X,Y coordinates to be able to provide a correct Z coordinate. I think it was okay in this instance because it was a small area and the relatively small amount of "tilt" correction would not appreciably affect the XY coordinates. But over a longer survey it might introduce a horizontal distortion if the tilt was significant. I'm thinking of some railway jobs coming up where we run a lot of precise vertical control with levels but don't necessarily collect a precise horizontal location for each benchmark.
I have used several survey adjustment software, like Star*Net, that allow me to hold any combination of X, Y, and Z for control coordinates, or better yet, specify different weights for each x, y, and z coordinate. Is there something I am missing in Cyclone that provides for a proper
But now I have a new project and it is in a 25 ft x 15 ft room, deep in the basement of a hospital, where among other things I need to provide a floor flatness and floor levelness report. This had to be a super tight survey to determine the intersection of the 3 axes of medical imaging machine within 3 mm, so I set 14 targets on the walls and 4 flat on the floor and then did 5 scans, one in the middle and one near each corner of the room ( I probably could have done it in 2 or 3 scans or possibly even one, but at 4 minutes per scan and a surveyors compulsion for redundancy, might as well get at least twice what I needed...) I then used a digital autolevel to run through the four targets on the floor to get true elevations on them to within about 2mm. I did this because knowing that the RTC360 does not have a DAC I wanted to have true level data in the end.
So I registered all my scans together and then exported the control vertices to a text file, and edited the text file and changed the Z coordinates of the targets to the leveled elevations. Then I imported the edited text file into Cylclone and registered the unified scan to the new leveled control. It worked out pretty good and I found that the "unleveled" scans had a tilt of almost 0.02 ft across the 25 ft length of the room, which I was able to correct with this method.
I am wondering if there is a better way to do this - I feel like my method was a bit bass ackwards in that I had to basically create fake X,Y coordinates to be able to provide a correct Z coordinate. I think it was okay in this instance because it was a small area and the relatively small amount of "tilt" correction would not appreciably affect the XY coordinates. But over a longer survey it might introduce a horizontal distortion if the tilt was significant. I'm thinking of some railway jobs coming up where we run a lot of precise vertical control with levels but don't necessarily collect a precise horizontal location for each benchmark.
I have used several survey adjustment software, like Star*Net, that allow me to hold any combination of X, Y, and Z for control coordinates, or better yet, specify different weights for each x, y, and z coordinate. Is there something I am missing in Cyclone that provides for a proper
-
- V.I.P Member
- Posts: 205
- Joined: Tue Mar 06, 2018 6:56 pm
- 6
- Full Name: Eric Guizzetti
- Company Details: Construction and Engineering
- Company Position Title: RealityCapture
- Country: USA
- Linkedin Profile: Yes
- Has thanked: 14 times
- Been thanked: 11 times
- Contact:
Re: RTC360 Inclinometer/IMU accuracy
are you talking about the view on the ipad or in Register 360 your seeing a tilt?spike3d wrote: ↑Mon May 13, 2019 9:51 pm Hi there,
I know some threads covered this briefly just after RTC360 launch. I was wondering if anyone verified tilt sensor/imu level accuracy on the RTC360. We're seeing better than 3' published in a registered data set of 300 scans of a large open building interior. We're not seeing any info exposed in Register 360 or Cyclone to be able to see the tilt sensor numbers or transformations. We will analyze the data more and come up with a more exact number but it appears better than 1'.
Has anyone analyzed it further?
Thanks
Spike
Eric Guizzetti
Long Beach, CA
www.dlgroup.com
www.eggreality.com
Insta: egg.reality
Linkedin - https://www.linkedin.com/in/ericguizzetti/
YouTube- https://www.youtube.com/c/ericguizzetti/
Long Beach, CA
www.dlgroup.com
www.eggreality.com
Insta: egg.reality
Linkedin - https://www.linkedin.com/in/ericguizzetti/
YouTube- https://www.youtube.com/c/ericguizzetti/
-
- V.I.P Member
- Posts: 205
- Joined: Tue Mar 06, 2018 6:56 pm
- 6
- Full Name: Eric Guizzetti
- Company Details: Construction and Engineering
- Company Position Title: RealityCapture
- Country: USA
- Linkedin Profile: Yes
- Has thanked: 14 times
- Been thanked: 11 times
- Contact:
Re: RTC360 Inclinometer/IMU accuracy
a quick fix for me on Field 360 has been reset the scanner, start a new scan position and it will fix alignmentpburrows145 wrote: ↑Tue May 14, 2019 10:10 amFW and SW improvements on the way to resolve the issue in a tiny fraction of cases. Also, a new tilt function has been added to the next release of Cyclone REGISTER 360 (as in FIELD 360) to quickly resolve this.Jamesrye wrote: ↑Tue May 14, 2019 8:43 am We use the RTC with targets because so many scans have come in "wonky". Have discussed the issue with Leica and SCCS who say Cyclone isn't reading the tilt sensor properly in some cases. We've been advised that a new version of Cyclone is on the way to resolve the issue. Currently, we are advised to press the button that looks like an 'R' in the visual alignment window to switch between levelled and unlevelled (or revert to the 3 point picks).
When doing a large building survey, I'd always do the stairwells with a P series scanner and/or shoot in control points with a total station.
If it is, then we are making the same procedural error.
Eric Guizzetti
Long Beach, CA
www.dlgroup.com
www.eggreality.com
Insta: egg.reality
Linkedin - https://www.linkedin.com/in/ericguizzetti/
YouTube- https://www.youtube.com/c/ericguizzetti/
Long Beach, CA
www.dlgroup.com
www.eggreality.com
Insta: egg.reality
Linkedin - https://www.linkedin.com/in/ericguizzetti/
YouTube- https://www.youtube.com/c/ericguizzetti/