Meeting Scope

• Autonomy has plans that they want to bring up

Comp Tasks

• LTE, ML, Airside autolanding worth 5 points

  • Progress not great, leads are burnt out. Cannot put in effort to get this done by competition.

  • Propose not doing auto-landing at competition, focus on task 1 stuff for competition.

    • Can continue work on them. Build support and infrastructure for future competitions.

• Task 1 worth 15 points

LTE

• Had 4 tasks:

  • 1st task complete - marked off

  • Get images from camera

  • Test image encoding

  • Need module to include that in our repository

  • Transmit data over network socket

• As it is - expected finish date is around middle of march. Members may be busy & have delays

• Big blocker is figuring out the image encoding that we want to choose

ML - Landing Pad Detection

• First round of data cleaning done. Don’t have enough images …

  • Need more images…

Airside

• Big blockers with integration tasks

• Detect target just merged into main

• as we start to integrate, there is a lot of effort that needs to be put into integrating them into main.

  • not hard to parallelize it into main

  • One person just needs to go in and grind out all of the bugs

• Geolocation will work perfectly if data from FC and location module is perfect

  • This can’t be guaranteed. Will have to see how it works with the error from ML model and FC

  • Cluster estimation will estimate where target is in that cluster

Task #1 system

• Set upper bound on # of laps the drone will run?

• Focus on MP for comp → dev up the long term stuff

Conclusion

Auto leads to work more on MP, start looking into more COTs sol’n and bring members up into it more. Continue dev on autoland, if both done then can benchmark.