Assessing the Positional Accuracy of the STA Explorer: A Field Validation Study

NOTE: This study was conducted on the STA explorer. The STA explorer HA was not tested, but is capable of centimetre level accuracy using PPK under the right conditions.

TL;DR

Across two consecutive days of static testing (n = 597 observations), the STA Explorer achieved a median horizontal accuracy (CEP50) of approximately 1.06 m, with 95% of positions within 1.8 m under open-sky conditions. Results were highly consistent between test sessions, demonstrating stable and repeatable metre-level performance when referenced against survey-grade RTK control.

Accurate and repeatable positional data underpins effective environmental monitoring, asset mapping, and operational reporting. To support transparency and continual improvement in our products, TerraLab regularly undertakes structured validation of our field devices under controlled conditions.

In February 2026, we conducted a multi-day field assessment to quantify the positional accuracy of the STA Explorer using a survey-grade RTK system as a reference. This article summarises the test design, analysis workflow, and key findings, with particular emphasis on the results of repeated static accuracy testing.

Test Overview

Equipment and Configuration

The following equipment and configuration were used:

• Device under test: STA Explorer
• Reference system: Emlid RX RTK receiver
• Mounting: STA Explorer mounted on the RX pole using a tablet mount (introducing a small physical offset)
• Corrections: AUSCORS network, Newhaven base station
• Collection software: ArcGIS Field Maps
• Processing software: ArcGIS Pro v3.5

The Emlid RX was configured to collect RTK-corrected positions in the GDA2020 reference frame. The STA Explorer recorded autonomous GNSS positions.

Weather conditions during testing consisted of negligible scattered cloud cover, providing near-ideal satellite visibility.

Test Design

Testing was conducted over three consecutive days and comprised three complementary components.

1. Static Survey Mark Test (19–20 February 2026)

A permanent survey marker was occupied on two consecutive days under identical field conditions.

• The RX was used to confirm the reference position of the marker.
• The STA Explorer remained stationary over the mark for approximately five minutes per session.
• Horizontal distances from each recorded position to the surveyed reference point were calculated.

These repeated static occupations provide the most robust assessment of intrinsic positional accuracy and repeatability.

2. Kinematic Polyline Test (18 February 2026)

An operator walked west along a footpath while recording:

• A continuous RTK polyline using the RX
• Simultaneous position records from the STA Explorer

Perpendicular offsets from the Explorer records to the reference polyline were calculated.

3. Kinematic Point Test (18 February 2026)

Discrete survey points were collected at 20–40 m intervals while moving east along the footpath.

• RTK points were collected with the RX
• The Explorer button was triggered at each location
• Distances between paired points were computed

Data Processing

All datasets were post-processed in ArcGIS Pro v3.5, including reprojection to a common spatial reference system and validation of reference frame consistency.

Distances between STA Explorer records and reference features were calculated using standard GIS proximity analysis methods. Distances represent planimetric separation between observed positions and surveyed truth.

All observations were retained for analysis, including initial start-up data.

Results

Data Quality Summary (Static Tests)

Across the two static test sessions, satellite geometry and signal quality remained consistently strong.

Parameter	19/02/2026	20/02/2026	Combined
Number of points	293	304	597
Mean satellites	32.5	33.95	33.22
Min HDOP	0.4	0.4	0.4
Max HDOP	0.5	0.5	0.5
Mean HDOP	0.41	0.40	0.41

These values indicate stable satellite availability and consistently favourable positioning geometry.

Static Test Results

The static survey mark tests provide the primary basis for accuracy assessment.

Radial Error Statistics (m)

Metric	19/02/2026	20/02/2026	Combined
Minimum	0.427	0.191	0.191
Maximum	1.851	2.780	2.780
Mean	1.038	1.105	1.072
CEP50	1.011	1.082	1.065
R68	1.213	1.285	1.242
R95	1.660	1.937	1.832

Where:

• CEP50 is the radius containing 50% of observations
• R68 is the 68% containment radius
• R95 is the 95% containment radius

Interpretation

Across both days::

• Median horizontal error (CEP50) was approximately 1.1 m
• 95% of observations were within approximately 1.8 m
• Maximum observed error was below 3.0 m
• Error distributions were consistent between sessions

The similarity of results between 19 and 20 February demonstrates strong short-term repeatability under stable field conditions.

The combined dataset (n = 597) provides a statistically robust estimate of typical device performance.

Kinematic Test Results (18 February 2026)

Kinematic testing was conducted on 18 February and is reported separately.

Test Type	Points	Mean Distance (m)	Min (m)	Max (m)
Kinematic Points	7	0.91	–	–
Kinematic Line	138	0.92	0.001	2.07

Kinematic performance was consistent with static results, with mean offsets below 1 m and similar upper bounds.

Discussion

Static Performance and Repeatability

Repeated static occupations on consecutive days produced highly consistent accuracy metrics, with CEP50 values of 1.01 m and 1.08 m, respectively. This indicates that observed errors are primarily attributable to inherent GNSS limitations rather than transient environmental effects. The combined results demonstrate stable metre-level accuracy under open-sky conditions.

Static vs Kinematic Performance

Static testing provides the most reliable measure of intrinsic positioning accuracy. Kinematic testing showed comparable performance, indicating that device filtering and motion do not introduce substantial additional error during normal field use.

Operational Implications

For environmental monitoring, compliance reporting, and asset mapping workflows, these results indicate that:

• Typical positional accuracy is approximately 1 m under open-sky conditions
• 95% of observations fall within 2 m
• Performance is repeatable across consecutive days
• No significant outlier behaviour was observed

This level of performance is consistent with high-quality mapping-grade GNSS systems.

Conclusions

This multi-day field validation study demonstrates that the STA Explorer provides stable, repeatable, and well-characterised positional data under favourable field conditions.

Key findings include:

• CEP50 of approximately 1.1 m in repeated static testing
• R95 of approximately 1.8 m across combined static datasets
• Consistent performance across consecutive days
• Comparable accuracy under kinematic conditions
• No evidence of significant systematic bias

These results provide confidence in the STA Explorer’s suitability for professional spatial data collection workflows where reliable, metre-level accuracy is required.

TerraLab will continue to undertake formal validation and calibration as part of its ongoing product development and quality assurance program.

If you’re interested in information about the STA logger, refer to this older post.