GUIDE-K35E
GUIDE-K35E tested with a full-band survey-grade antenna in a vehicle-based Ground Truth field test.
View GUIDE-K35E product page →Ground Truth Field Test · 01
GUIDE-K35E GNSS Ground Truth Field Test: Vehicle Route Performance
This Ground Truth field test evaluates the GUIDE-K35E GNSS module in a vehicle-based route dataset. The test uses a full-band survey-grade antenna and documents measured positioning performance against the defined ground-truth reference.
Test Overview
Test conditions & dataset
Basic Test Information
Test time
2026-06-08
Test platform
Car
Antenna
Full-band survey-grade antenna
Test sample
GUIDE-K35E
Dataset Summary
Output rate
10
Hz
Epochs
69,901
Total distance
58.8
km
Duration
01:56:30
Route Overview
Test Samples
What was tested
Overall Performance
Measured full-route performance
Full-run results retain normal, degraded, and dead-reckoning intervals so that measured accuracy can be read alongside solution quality.
Full-run results
GUIDE-K35EHorizontal RMS
2.688
m
Vertical RMS
0.808
m
CEP50
0.015
m
R95
4.825
m
Position Quality & Accuracy Distribution
Two aligned cumulative bars over the full run. Matching colours connect each accuracy band to its corresponding position-quality tier. Hover a segment for detail.
Position Quality
- RTK Fixed 57198 ep 81.827%
- RTK Float 10367 ep 14.831%
- DGNSS 29 ep 0.041%
- Single 70 ep 0.1%
- DR 2237 ep 3.2%
Accuracy Range
- <10 cm 57338 ep 82.027%
- 10 cm–1 m 1872 ep 2.678%
- 1–3 m 4371 ep 6.253%
- 3–10 m 5294 ep 7.574%
- >10 m 1026 ep 1.468%
Overall Error Timeline
Move across the chart to inspect horizontal and vertical error at the same time.
Loading interactive error timeline…
Horizontal error magnitude
Vertical error, signed
Solution quality status
Scenario Results
Per-environment performance
Each scenario shows the desensitized trajectory video with in-video error readout, plus the same standardized statistics.
Accuracy Range
<10 cm
10 cm–1 m
1–3 m
3–10 m
>10 m
Position Quality
RTK Fixed
RTK Float
DGNSS
Single
DR
01
Highway
Open-road highway conditions produced stable RTK-fixed positioning with a CEP50 of 0.016 m and negligible quality transitions across the selected 17.5 km open-sky segment.
Duration
00:12:10.0
Distance
17.5 km
Epochs
7301
Highway
0:00 / 0:00
Core Accuracy Metrics
Horizontal RMS
0.022
m
Vertical RMS
0.018
m
CEP50
0.016
m
R95
0.039
m
02
Residential
Low-rise residential streets maintained essentially full RTK-fixed coverage at centimetre level (CEP50 1.3 cm), showing no measurable degradation from mild roadside obstruction.
Duration
00:10:58.0
Distance
2.8 km
Epochs
6581
Residential
0:00 / 0:00
Core Accuracy Metrics
Horizontal RMS
0.021
m
Vertical RMS
0.026
m
CEP50
0.013
m
R95
0.037
m
03
Urban Canyon
Dense urban streets with intermittent building obstruction maintained predominantly RTK-fixed positioning, while brief RTK-float transitions appeared in reduced-sky-view sections.
Duration
00:17:36.0
Distance
3.6 km
Epochs
10561
Urban Canyon
0:00 / 0:00
Core Accuracy Metrics
Horizontal RMS
0.023
m
Vertical RMS
0.042
m
CEP50
0.014
m
R95
0.045
m
04
Tunnel
Tunnel passage caused a sustained GNSS-denied interval. Dead reckoning accounted for 85.5% of the selected window, with a horizontal CEP50 of 1.900 m before satellite positioning was re-acquired.
Duration
00:01:47.9
Distance
2.4 km
Epochs
1080
Tunnel
0:00 / 0:00
Core Accuracy Metrics
Horizontal RMS
2.178
m
Vertical RMS
0.872
m
CEP50
1.9
m
R95
3.727
m
TEST NOTES
Reading this test
Open-road performance is exactly what you'd want, and not much more to say.
On highway and open segments the receiver stayed RTK Fixed almost continuously and tracked the ground truth at centimetre level — steady and uneventful, which is the point.
The dense-city result beat our expectations.
Residential streets and the CBD canyon held fixed nearly as well as open road — a glimpse of what a dual-band engine and a survey-grade antenna can do together. We'll take this further at the product level, starting with SCOUT PRO, to see how it holds up in practice.
The tunnel is a dead-reckoning story, and we'll be candid about it.
With no sky the solution coasts on MEMS inertial dead-reckoning — error drifts gradually rather than jumping, and snaps back to a fixed solution quickly on exit. It coasted well this run, but the IMU is uncalibrated (no temperature or bias compensation) and parts vary unit to unit, so it won't be this clean every time. Treat these values as coasting behaviour, not GNSS accuracy.
EXPLORE THE PLATFORM