I have two recent events that has diminished our confidence in our RS3 receivers. I’m hoping that you might have a plausible answer for our experience. We are seeing a drift of about 0.50’ from start of day to end of day… or day-to-day.
First experience:
San Antonio Texas, RS3, corrections supplied by PointOne TruRTK via internal cellular modem. Start of day localized on five control points supplied by the project surveyor. As-builted around 1,000 points during the day. Over the course of the day the operator kept indicating that certain features were not lining up as expected and indicated about a 0.50’ (roughly) horizontal distance from the expected point location. Initially, the operator kept working to as-built the locations, thinking that the construction is actually off a little. At the end of the day, the operator checked “staked” to one of the localization control points and found a horizontal error of about 0.55 feet and a vertical error of about 0.40 feet. We could not find reasoning, or a theoretical cause so we abandon our full day of data and re-collected the next day with our traditional Topcon RTK receivers.
Second experience:
El Paso Texas, RS3 (different unit from one above), corrections supplied by PointOne TruRTK via internal cellular modem. The point data used a previously localized coordinate system saved in Flow. Operator (different from above) staked and marked about 500 points before realizing that the staked locations did not match staked locations from a previous site visit about 3-4 weeks prior. The horizontal error was about 0.50 feet from the previous session, vertical error was not checked by the operator. We are unsure if the previous staking session was inaccurate, or the new staking session was inaccurate and now we have to plan a new site visit to rectify the situation with trusted equipment.
Both experiences happened last week. Two different receivers set up exactly the same, 33 Beta 1 firmware, two different operators, two unique locations, approximately the same apparent error of around 0.50’. Other details include, tilt compensation generally not used for collecting data, but is used for staking data. Localization residuals are tight and without abnormalities. My gut feeling is that the tilt compensation feature, when turned off still effects the measurement and over time the tilt-compensation calibration goes out of range and unintentionally effects the measurements. I’m hoping that you can help us to understand the inconsistencies with our new RS3 units. We won’t use them again until we figure this out. Everything else about the receivers and software is great. Thanks -Jeff
With such extensive work on what appears to be a very active site, I’m curious why your workflow doesn’t use a project benchmark on site on which to set up your base rather than use a network RTK service? That would seem to eliminate some factors in your troubleshooting.
The 0.50 ft difference is suspicious. Did I understand correctly that the error is systemic? Does that mean that all the points have a difference of 0.50ft when staked out or recollected at a later time or day?
Do you remember how long the baseline was? Could you also share the CSV from Flow? Since it contains sensitive information, you can email it to me at support@emlid.com.
For San Antonio, we are unsure as to when the drift started to occur. We think the day started OK because the constructed features were aligning at the locations that were expected, however at some point the constructed features appeared to be slightly out of alignment and eventually it was decided that the error was approximately 0.50’. When the operator checked back into a localization point he noted the error to be 0.55’H / 0.40’V from the position at the beginning of the day. I don’t have a specific baseline distance from the correction service base, however the area is well covered.
For the El Paso project, the operator only noticed the error in position at the end of the day, when compared to the staked location of a previous work session using the same RS3 receiver. The correction service base is located less than 10km from the project. Thank you for looking into this issue.
The point of trying the Emlid receivers was to get away from setting a base and traveling with all of the extra equipment. We were hoping to be able to fly between projects instead of driving, which seemed like a reasonable expectation given the features of the RS3.
You are flying to work on a project and don’t want to include an extra receiver and mount. That must mean you don’t travel with any backup gear either?
Hi Christian-
No, we have never checked the ionospheric disturbances and would not know what to do with an ionospheric report. We don’t have any photos of the localization points, however the San Antonio site is clear of any trees and has underground electrical. The El Paso site has some trees and other obstructions, similar to almost every construction site. The issue is that this same anomaly occurred on two different RS3 receivers on different days at different locations with different operators. I am doubtful that there is an environmental factor that is to blame. -Jeff
