We’re UAV GNSS (uav-gnss.com), a Septentrio receiver maker, and one of the first decisions every AgOpenGPS build hits is where the corrections come from. Here’s how we tell customers to choose between survey-in and a fixed base.
Survey-in — the receiver averages its own position over time (typically 60-300 s) until it hits your accuracy target (e.g. 2 cm). No surveying, no setup, works anywhere. The catch: the averaged position is only as good as the RF environment during the averaging window. If a truck parks beside the antenna mid-survey, or a cell tower is close by, that bias gets baked into the base position — and your whole field is offset by it.
Fixed base — you enter known coordinates (surveyed point, known marker, or a network RTK position you trust). This is the only way to work in a UTM/project coordinate system that matches an existing map, or to get repeatable absolute accuracy across multiple days.
Practical guidance:
- Single-day job on a small field, no existing grid → survey-in is fastest and fine.
- Multi-day job where rows must line up with last week’s, or you’re merging data with a surveyor → fixed base with surveyed coordinates.
- Either way: stable tripod, clear sky view, away from metal. A base fighting multipath degrades every rover using it.
- Re-survey if you move the base more than a few cm — a shifted base means a shifted field.
Rover side: even with a perfect base, the rover’s front end must survive the RF environment (power lines, electric fences, 4G towers). Receivers with AIM+ anti-jamming (~60 dB) hold fixed where consumer modules (~25 dB) drop out.
Wiring and parameter guides: GitHub - uavgnss6-bot/septentrio-gnss-integration-guide: Septentrio GNSS receiver integration guide for ArduPilot and PX4 — wiring, configuration, RTK setup, and SBF data parsing. · GitHub. Do you run survey-in or a fixed base — and how do you establish the base coordinates?