Data

Flight Data Import

Bring your flight computer log in as a recorded flight.

At minimum, the importer needs a time column and either altitude or pressure. Map those columns, check the detected events, and the log can be saved as a recorded flight. If it also includes acceleration or GPS data, ZenRockets uses those channels to improve the reconstruction. Other sensor channels remain available in the charts.

Import a flight#

  1. Open Flights from the dashboard, select Import Flight, and upload a CSV.
  2. Map the time column and either altitude or pressure. Set the unit for each mapped column.
  3. Review the reconstruction method. If the log has no horizontal source, choose a wind profile and airframe coupling for the synthetic drift model.
  4. Review the detected liftoff, burnout, apogee and landing times. Adjust them or add drogue and main deployment events when needed.
  5. Name the flight, optionally link its rocket and launch details, then import it.

The saved flight opens with a 3D trajectory, charts for mapped channels, and the same comparison tools as a simulation.

Map your columns#

The mapping step pairs each CSV column with what it means. Each mapping connects one CSV column to the field it represents in the flight, with the column's unit set between them. Time is always the first row (required, and locked to its field), and Add Mapping adds as many more as the log has channels. Once a CSV column is mapped, it is unavailable in the other rows. The unit and field choices also stay compatible, so a column measured in feet can only be mapped to a length field.

The file's first rows are previewed above the mappings, and a header row is detected automatically — its names become the source-column labels.

Target fields are grouped by what they describe:

  • Position — altitude, x/y position, latitude and longitude
  • Velocity — total speed or per-axis velocity
  • Acceleration — the acceleration channel used in reconstruction
  • Raw sensors — raw accelerometer axes
  • Rotation and rates — roll/pitch/yaw angles, angular velocity and angular acceleration
  • Atmosphere — pressure and temperature

To continue, the wizard needs time plus a vertical source (altitude or pressure), and the time column's values have to increase; everything else is optional.

What each column adds#

Altitude or pressure

Altitude or pressure is required because the importer needs a source for vertical position. Altitude provides it directly. Pressure provides it indirectly by converting each reading through a standard-atmosphere model. Invalid pressure readings are skipped.

Acceleration

When acceleration is mapped alongside altitude or pressure, the importer fuses the channels to capture fast changes during boost and burnout. If the acceleration data appears wrong or inconsistent, the importer falls back to altitude or pressure alone.

GPS or positions

Latitude/longitude give the trajectory its horizontal shape; measured x/y positions or velocities are used directly when the log has them. Without either, a wind-drift model fills in a plausible descent path.

Everything else

The remaining mapped sensor channels stay available in the charts but are not used to reconstruct the trajectory.

Trajectory reconstruction in depth#

The importer reconstructs the vertical motion and horizontal track separately, based on the columns you mapped. It then places both on the same timestamps so the trajectory, charts and events stay aligned.

The vertical channel#

The vertical reconstruction starts with altitude. If the log contains pressure instead, each pressure sample is converted to altitude using the seven-layer International Standard Atmosphere model. Zero or negative pressure readings are skipped so one bad row cannot corrupt the profile.

How velocity and acceleration are recovered depends on whether you mapped an accelerometer:

  • Altitude only. The profile is Savitzky–Golay smoothed, then analytically differentiated to recover vertical velocity and acceleration. This works well through descent and around apogee, although a barometer alone can soften fast transitions such as burnout.
  • Altitude plus acceleration. A Kalman filter combines the sources. The accelerometer bias is estimated from the pad period, gravity is removed, and the corrected signal is integrated as a cross-check against the barometric profile. If the sources disagree — for example, because the wrong axis or unit was mapped — the importer discards the accelerometer and falls back to altitude-only reconstruction. When the acceleration and barometric data agree, acceleration is used through boost and coast. Near apogee, the reconstruction relies on the barometric altitude for descent.

Mapped vertical velocity remains chartable, but it is not used to reconstruct the vertical channel. Flight computers often report heavily filtered velocity that lags fast phases, so reconstructed velocity comes from altitude and optional acceleration instead.

The horizontal track#

For the horizontal track, the importer uses the available sources in the following order:

  • Measured x/y or vx/vy. If the log contains x/y position or velocity, the importer uses it to build the horizontal track.
  • GPS latitude/longitude. GPS latitude and longitude provide the ground track relative to the launch pad. Invalid or stale fixes are ignored.
  • No horizontal source. The selected wind profile and airframe coupling generate a synthetic drift path against the vertical profile. The effect is concentrated during descent, where wind contributes most of the horizontal motion.

One time grid#

Every channel is resampled onto the vertical profile's timestamps. Horizontal velocity is derived from the smoothed track when it was not measured directly.

Short gaps in GPS data are estimated to keep the track continuous. Once valid data returns, the importer resumes using it.

Event detection and review#

The importer automatically detects liftoff, burnout, apogee and landing from the reconstructed vertical data. Burnout uses the best available acceleration signal, preferring a consistent measured channel and falling back to reconstructed acceleration when necessary.

The wizard shows those events before saving, each with its interpolated altitude and vertical speed. Detected events keep their type, but their times are editable. Add Event lets you add your own: the four core events plus Drogue Deploy and Main Deploy, each usable once. Parachute deployments are never inferred automatically, so if the flight had them, this is where you add them. Any event can be removed.