Accuracy comes from the whole workflow. The drone, the GNSS data, the overlap, the ground control, the processing, and the QA all have to line up. The weakest link usually sets your accuracy ceiling.
For many practical survey drone jobs, teams target around a tenth of a foot. That is a useful benchmark, but only when the project actually needs it and the field plan is built to support it.
Questions to answer upfront
- What horizontal and vertical accuracy does the deliverable actually need?
- Is the site open enough for photogrammetry, or is lidar more appropriate?
- Is your aircraft RTK/PPK enabled?
- How much ground control or check data will be collected?
- How much field time can you afford without sacrificing the result or return on investment?
Know your required accuracy up front: field workflows can differ drastically in flight time and overall field time depending on your accuracy tolerance and the GPS capabilities of your drone/sensor.
Out with the Old and in with the New (drone hardware)
Modern RTK and PPK-capable systems make good accuracy much easier than older aircraft. That does not mean ground data stops mattering. It means the control strategy can be more efficient. Older aircraft without strong onboard positioning usually require more conservative procedures and tighter GCP planning to get to the same result.
What this means in the field
If the project truly needs high-confidence survey-grade output, plan the flight overlap, altitude, control distribution, and checkpoint strategy around that need. If the project only needs one-foot contours or a planning-grade site model, plan around that need and save some flight and field time. One of the biggest missteps we see here are flight parameters aimed at "the best accuracy possible", which increases field time, operational complexity, data footprint, and processing time, unnecessarily.