Leica RCT360 level check

Discuss all Leica Geosystems related issues here.
Post Reply
User avatar
topogeo
I have made 90-100 posts
I have made 90-100 posts
Posts: 91
Joined: Fri May 02, 2014 1:38 pm
Full Name: piercarlo roasio
Company Details: studio roasio
Company Position Title: titolare
Country: italy
Linkedin Profile: No
Has thanked: 1 time
Been thanked: 8 times

Leica RCT360 level check

Post by topogeo »

I made several tests to verify the reliability of the horizontally of a group of RTC360 scans (which as we know is devoid of a real compensator but uses only the IMU).
I checked on real works done outside: 30-40 scans in areas that are not always optimal (grass on the ground, some trees, buildings not arranged in an optimal way .. as it happens in reality). Distance between targets within 100-150 meters (but in some jobs even 300 meters)
Registration made with Rgister360
The scans are at 20-25 meters maximum and I have always removed trees and grass in the visual alignment (... but without wasting too much time)
The links between scans have 2-4 mm offsets.
I almost always have closed loops with global errors between 1 and 5 mm.
In Register360 I measured the distance and height difference between the B / W targets (automatically recognized in the import phase) BEFORE applying the control points.
I compared distances and height differences between targets in Register360 and the same targets detected with GNSS (Leica GS12) or total station (Leica MS60 1 ").
The results were very surprising!
The differences are 1-2 cm every 100 meters with the targets detected with the total station (therefore with a precision of 2-5 mm)
They become 2-4 cm every 100 meters if the targets have been detected with GNSS. In these cases, however, the topographical survey is certainly less precise because the GNSS is very affected by obstacles and it was not always possible to place the targets in very open areas.
I checked about 12 jobs ...
Result: the scanning groups with RTC360 are leveled with an accuracy better than 2-3 cm every 100 m!
So the many fears about the lack of compensator seem ... not justified.
Has anyone done similar checks?

Piercarlo
Leandre Robitaille
I have made 70-80 posts
I have made 70-80 posts
Posts: 73
Joined: Sat Aug 03, 2019 1:53 am
Full Name: Leandre Robitaille
Company Details: Cima+
Company Position Title: Civil Technician
Country: Canada
Has thanked: 2 times
Been thanked: 11 times

Re: Leica RCT360 level check

Post by Leandre Robitaille »

Interesting test you did there! The lack of compensator is the main reason why we didnt buy the rtc360. The issue I have with your tests is that there is an actual probability that you would be perfectly leveled at a godlike level of leveling, its all a probability and we are looking for reliability. To be more conclusive you would have to run more tests, all you need is 1 bad result and there goes your reliability. Would be also interested to see how this compares to a faro s70 scanner. Just my 2cents

According to the specs you can get up to 9cm error at 100m. Vs a compensator such as a faro that is up to 2.6cm at 100m
jamesworrell
V.I.P Member
V.I.P Member
Posts: 367
Joined: Mon Jun 16, 2014 1:45 pm
Full Name: James Worrell
Company Details: Bennett and Francis
Company Position Title: Director
Country: Australia
Linkedin Profile: Yes
Location: Brisbane, Queensland, Australia
Been thanked: 22 times
Contact:

Re: Leica RCT360 level check

Post by jamesworrell »

We regularly see similar results - bundles with loops etc do tend to perform quite well.

I tend to call it GPS-grade. And indeed a few gps marks past the extents of a survey will generally be enough for most data this might be presented to say 2-decimals.

A total station certainly gives a level of comfort - and a long linear job - say a roadway, I'd suggest would need something be it TS or GPS.
mappilots
I have made <0 posts
I have made <0 posts
Posts: 6
Joined: Thu Aug 27, 2020 9:17 am
Full Name: Daniel Michalczyszyn
Company Details: Mappilots
Company Position Title: Pilot
Country: Norway
Skype Name: fugazii01
Linkedin Profile: Yes

Re: Leica RCT360 level check

Post by mappilots »

Leandre Robitaille wrote: Sun Sep 13, 2020 2:21 am Interesting test you did there! The lack of compensator is the main reason why we didnt buy the rtc360. The issue I have with your tests is that there is an actual probability that you would be perfectly leveled at a godlike level of leveling, its all a probability and we are looking for reliability. To be more conclusive you would have to run more tests, all you need is 1 bad result and there goes your reliability. Would be also interested to see how this compares to a faro s70 scanner. Just my 2cents

According to the specs you can get up to 9cm error at 100m. Vs a compensator such as a faro that is up to 2.6cm at 100m
Dear Piercarlo Roasio,
as Leandre Robitaille mentioned, one area, one project does not reflect the situation. We are currently dealing with a very large project. Parallel week after week, we have projects like 22,000 square meters + and so for 4 years.
We were delighted with the speed of the RTC360 scanner, and the ease of use, after some time we realized that the registration in Register 360 is not always correct and the green light can be wrong.
That's why we no longer rely only on Register 360 for complicated places (Register 360 is mostly fine with cubature buildings), Cyclone Basik provides proven results. It's not that the Register 360 is bad, but we had bucket of cold water, and we already approach it carefully, we give many more reference points than we assumed when we were buying RTC360, we clean the point cloud more thoroughly than we assumed before. After such experiences, we are very careful with Register 360.
And now the matter of the compensator, in scanners with a compensator you get an additional tool that is your voice of conscience.
If station, after performing the cloud to cloud registration has changed the original compensator value, it is a sign for you that, for example,
- noise caused 2 stations register badly,
- 2 stations merged into one,
- noise caused vertical wobble.
The point here is not only that the compensator will do everything for you, but it gives you information that during the alignment this station was vertically breached. You don't need to look for such situations now throughout the project, they will be highlighted in the results.
You feel more certain that this project will not come back to you from the architect, and you will be shamed.
Such a thing, looking ahead in combination with RTC 360 would tell you, ok...here cloud 2 cloud pushes my compensator out of balance, time to target some station, you do not have to rely only on intuitions and experience.

Kind Regards
Daniel Michalczyszyn
Anko AS
User avatar
topogeo
I have made 90-100 posts
I have made 90-100 posts
Posts: 91
Joined: Fri May 02, 2014 1:38 pm
Full Name: piercarlo roasio
Company Details: studio roasio
Company Position Title: titolare
Country: italy
Linkedin Profile: No
Has thanked: 1 time
Been thanked: 8 times

Re: Leica RCT360 level check

Post by topogeo »

jamesworrell wrote: Sun Sep 13, 2020 3:17 am We regularly see similar results - bundles with loops etc do tend to perform quite well.

I tend to call it GPS-grade. And indeed a few gps marks past the extents of a survey will generally be enough for most data this might be presented to say 2-decimals.

A total station certainly gives a level of comfort - and a long linear job - say a roadway, I'd suggest would need something be it TS or GPS.
I really like the term gps grade! :D :D
In fact, the accuracy of an RTC360 survey depends on the topographical quality of the control points.
I always use a lot of them (1-2 per scan) .. usually GNSS.
My tests say that the RTC360 level is better than it should be .. but .. I always prefer to have control points (also to verify the merging of the scans .. and to georeference the work).
I have been a topographer since 1989 and I have been using the laser for a few years. I always prefer to be able to check my work (.. and I know that the only safe verification is with the total station .. while the GNSS is less reliable ..)

Piercarlo
User avatar
landmeterbeuckx
V.I.P Member
V.I.P Member
Posts: 698
Joined: Tue May 01, 2012 5:19 pm
Full Name: Lieven Beuckx
Company Details: Studiebureau Beuckx
Company Position Title: Owner
Country: Belgium
Linkedin Profile: Yes
Has thanked: 46 times
Been thanked: 122 times

Re: Leica RCT360 level check

Post by landmeterbeuckx »

topogeo wrote: Mon Sep 14, 2020 11:14 am
jamesworrell wrote: Sun Sep 13, 2020 3:17 am We regularly see similar results - bundles with loops etc do tend to perform quite well.

I tend to call it GPS-grade. And indeed a few gps marks past the extents of a survey will generally be enough for most data this might be presented to say 2-decimals.

A total station certainly gives a level of comfort - and a long linear job - say a roadway, I'd suggest would need something be it TS or GPS.
I really like the term gps grade! :D :D
In fact, the accuracy of an RTC360 survey depends on the topographical quality of the control points.
I always use a lot of them (1-2 per scan) .. usually GNSS.
My tests say that the RTC360 level is better than it should be .. but .. I always prefer to have control points (also to verify the merging of the scans .. and to georeference the work).
I have been a topographer since 1989 and I have been using the laser for a few years. I always prefer to be able to check my work (.. and I know that the only safe verification is with the total station .. while the GNSS is less reliable ..)

Piercarlo
1-2 per scan? I do like 6 in total, 2 at the start, 2 in the middle and 2 at the end normally. If a larger project or other type of constellation i add more.

Of course gps, if you were to use total station for all these points you'd better measure it the classic way ;)
LSBbvba
Surveying services - 3D Laserscanning
Tel : +32477753126
www.lsbbvba.be
[email protected]
User avatar
topogeo
I have made 90-100 posts
I have made 90-100 posts
Posts: 91
Joined: Fri May 02, 2014 1:38 pm
Full Name: piercarlo roasio
Company Details: studio roasio
Company Position Title: titolare
Country: italy
Linkedin Profile: No
Has thanked: 1 time
Been thanked: 8 times

Re: Leica RCT360 level check

Post by topogeo »

[/quote]
1-2 per scan? I do like 6 in total, 2 at the start, 2 in the middle and 2 at the end normally. If a larger project or other type of constellation i add more.
...
[/quote]

1-2 for each single scan> 20 scans = 20-30 control points retrieved with GNSS (obviously it depends on the site: in a historic center with narrow streets I can put less. I always look for areas where I have good air the GNSS: I don't trust points with 3D quality worse than 2.5 cm- Spx GNSS points with 3-4 cm quality have errors, especially at altitude, even of 5-7 cm) If I have some cornerstones outside the large relief site for example 150 meters, an error of 2-3 cm I consider it acceptable (..gps grade accuracy)
I also make sure that the targets are well distributed all around: I don't want targets on a line (... the work could rotate along that axis)

Piercarlo
User avatar
landmeterbeuckx
V.I.P Member
V.I.P Member
Posts: 698
Joined: Tue May 01, 2012 5:19 pm
Full Name: Lieven Beuckx
Company Details: Studiebureau Beuckx
Company Position Title: Owner
Country: Belgium
Linkedin Profile: Yes
Has thanked: 46 times
Been thanked: 122 times

Re: Leica RCT360 level check

Post by landmeterbeuckx »

With that many points which vary in height constantly, isn't tying everything together is more difficult?

The least points the better i would say. i don't know the process with Leica of course but how doe sthe software decide which is best? Can you overrule specific data?
LSBbvba
Surveying services - 3D Laserscanning
Tel : +32477753126
www.lsbbvba.be
[email protected]
User avatar
topogeo
I have made 90-100 posts
I have made 90-100 posts
Posts: 91
Joined: Fri May 02, 2014 1:38 pm
Full Name: piercarlo roasio
Company Details: studio roasio
Company Position Title: titolare
Country: italy
Linkedin Profile: No
Has thanked: 1 time
Been thanked: 8 times

Re: Leica RCT360 level check

Post by topogeo »

I put a lot of points because ... it costs me nothing :)
We always go to make the surveys in two operators. One to the scanner and the other with the GPS puts the targets.
I prefer to have many and discard unreliable points.
The GPS provides a quality index that is NOT always reliable (... it is a long speech that surveyors know well: the GPS ESTIMATES the position of a point and like every estimate it has an error that is estimated ... but it can be inaccurate ( and the manufacturers cheat a little in the filters .. in order to show better results than the real ones). Only the total station MEASURES a point (always with errors .. but much more limited and .. known). But it takes more time and therefore we use it only for surveys that must be very precise or where the GPS cannot work (for example streets of historic centers).
In Leica Register360 I insert all the tergats then I discard those that are not reliable. For example if I have 3-4 targets in a courtyard ... all with quality (GPS) 3 cm ... but then when I block the connections between the scans and apply the control (and then Register does a rigid roto-translation) ... on one c 'is a much higher deviation .. it is LIKELY that the GPS survey of that point is inaccurate .. so I remove the point ..
In Register360 this operation is very fast. Many tests can be done in seconds.
Piercarlo
virtuoso
I have made 20-30 posts
I have made 20-30 posts
Posts: 20
Joined: Mon May 25, 2020 12:56 pm
Full Name: Antonio
Company Details: Mck
Company Position Title: BIM engineer
Country: Cro
Has thanked: 2 times

Re: Leica RCT360 level check

Post by virtuoso »

Hi guys, I'm currently in progress of learning georeferencing pointcloud in Register 360 (Before, investors weren't interested in georeferencing pointcloud, they only wanted 3d model so they can get accurate plans of building) as our current investor want's everything georeferenced. Atm I have job with around 300 scans (multiple bundles joined together). Tru slice seems good and after applying control, everything rotates as it should. The area that is scan is around 300 m in length, and somewhere i get difference around 5-10 cm. Worth noting that we don't have inhouse people that do georeferencing, so we usually contact another firm to use drone + put targets and measure those points. Some of those points are not black and white (I used virtual targets on those) and are located on grass where terrain is not straight. Would this be relatively OK georeferenced? Asking those who have lots of experience. Also how accurate could be RTC 360 vs drone(not LiDAR but taking photos and those are converted in pointcloud)?

Best Regards
Post Reply

Return to “Leica Geosystems”