https://en.wikipedia.org/wiki/GPS/INS

GPS/INS is commonly used on aircraft for navigation purposes.
Using GPS/INS allows for smoother position and velocity estimates that can be
provided at a sampling rate faster than the GPS receiver. This also allows for
accurate estimation of the aircraft attitude (roll, pitch, and yaw) angles.

A post from Dhayner to
https://ozreport.com/forum/viewtopic.php?t=56724

Track Logs: GPS vs. INS and Blending, 16 Jul
2018 8:46 pm

As I recall, the Oudie boasts of containing 3-axis gyros and 3-axis
accelerometers as part of an IMU. The correcting altimeter should be pressure
based to eliminate and dependence on the GPS. The track log looks like many INS
only tracks I’ve seen – not random, just a drift component due to biases and
drifts native to all inertial sensors.

What I suspect is that the INS track is used to “filter” the GPS track to remove
the GPS jumps and dither. Also, if the GPS data quality falls below some level,
the INS track is more heavily weighted than the GPS track. Once a good, high
quality GPS data track is re-established, the INS is reset to the GPS track.

The jump, or correction we see in the data near the goal may be Oudie software
deciding the GPS data is of sufficient quality to start to weight it more
heavily (and the INS track less).

The developers need to disclose this in detail and provide some sort of visual
indicator (in all screens) that the GPS data is being ignored or de-emphasized.
Unfortunately, depending on the algorithm details, the developers may not even
know. There are many pre-packaged INS/GPS blending apps available that are
extremely sophisticated, but little details are provided on what is going on
deep in the guts of the algorithms.

I’ve also seen many “professional products” developed by teams essentially
ignorant of the details of the Kalman or Particle filters employed and how
statistical measures of signal quality are used.

As an example of my concern of the Oudie developers understanding of how to use
inertial sensors, they claim on their Web site:

Exceptionally accurate Vario (Gyro assisted) for the quickest climb rate

Gyros measure angular rate which has little to do with moving up or down in a
gravitational field. Accelerometers, corrected for sensing angle by the gyros,
could assist. But it is the accelerometer data that is measuring “up” or “down”
accelerations, not the gyros.

The performance of the small, low cost MEMS gyroscopes and accelerometers
employed in these products have drift rates and uncorrected biases far beyond
practical use for INS only navigation for hang gliding or paragliding competitions. Given these
errors and several other reasons, I suspect CIVL will need to specify that GPS
only track logs will be the only accepted track.