Radar engineer upskill

Hello all,

I’m a radar signal processing engineer (point clouds, spectrum analysis, lots of legacy debugging) and want to move into applied ML for robotics. I have a masters in robotics and AI.

I’ve got solid math + sensor data experience, and access to real data plus an internal repo with ML projects.

My main question: is it worth spending time re-implementing those ML algorithms myself plus doing side projects, or is not worth it.

I can dedicate 2 hours a day for the projects. I am very serious about leaving, but i lack direction.

Would you:

• Stay and build projects on the side?
• Try to pivot internally?
• Or consider something like try to do research with a professor?

If you’ve made a similar move, what actually helped you break in?