Large covariances on backpack 2

[mcorah]

Backpack 2 seems to be reporting large covariances

• Refer to bags: Are the GNSS1/2 fix_info topics reporting enough satellites (15+)
• Look at gnss1/2 odometry topics. Diagonal of covariance are variances in meters-squared. How large are the standard deviations in meters? (should be less than 10cm when operating as it should)
• Could test live as well
• Try also comparing to times when a GPS was performing well... Micah seems to remember seeing 10cm or better accuracy in past

@hannah-yes These would also be good things to look at plotting!

Note! This is the backpack with different antennas! Do these antennas just perform worse? (would be good to compare to other backpacks).

[mcorah]

This seems to be a matter of improper EKF configuration. M3 seems to have been configured correctly and provides low covariance on /nav/odom/. Backpacks (backpack 3 in this case) are providing low covariance on gnss*/odom but 1600 on /nav/odom/.

@seungchan-kim Would you be able to look into this?