eskf with sensor concept, less states, and latest fixes from dev/auv-navigation-eskf

[AhmedBorch]

Hi, I did the following modifications:

• added the sensor concept requirements.
• removed gravity from the states because it doesn't change and computation would be faster without it -> reduced matrices sizes
• add the modifications from Talha's branch: naming convention, latest fixes, use of utils.
• changed the params (IMU is not rotated in the simulation)
• I had to add a .tpp file for the function template
• I used NDEBUG for publishing NIS, so it will only be published when building with debug flag.

btw I didn't branch from Talha's because I had already made lots of modifications, it was easier to implement the latest missing fixes from his branch in here, I hope this is okay:)

P.S:

• R is 0 (maybe I should add some values in the simulation)
• I will make a separate file to calculate errors and nees
• the S is very big in the beginning but then gets smaller, I will need to investigate this more (maybe have a big initial covariance)

[Andeshog]

Your branch is behind the main branch, and that is causing merge conflicts. Update that and make sure all tests pass 😄 Also, you could still change the target branch to Talha's branch so the PR can be continued there, but it probably doesnt matter if this is up to date.