Reddit reviews Feedback Control of Dynamic Systems (7th Edition)

So, robotics/mechatronics is not really my area of expertise, but what I think can be helpful is distinguishing between the different skillsets involved in the area. Companies don't really recruit just general "roboticists" or mechatronics people. Sure, you'll have a leg up if you've got some cross-disciplinary familiarity, but the closer you are to the cutting edge, the more it matters that you're really, really good at one particular thing rather than okay at all the things involved.

Again, this is not really my field, but a few of the things I'd look to focus on are:

Sensing - Almost every sensor system out there has notable room for improvement, whether that's the hardware being too expensive, too big, too low in resolution, etc. or the signal processing algorithm being too slow, too narrowly effective, or too complicated for anyone but the PhD who wrote it to implement it.

Dynamics & Controls - This is the real meat and potatoes of robotics, in my opinion. There's no one-size fits all controller, and getting this right is the make or break area for most robotics systems. Incredibly deep field, and if you're looking to do actual robotics R&D, this will be where you need to end up.

Mechanical Design - Useful, but not really a driver for most robots. Most are well within the stress/strain limitations of their materials and aren't driving for ultra lightweight construction or needing exceptionally capable on-board power generation or heat management. Major caveat - if you're doing something dynamically complex and/or aerodynamic, the line between the mechanical design and the controls aspects evaporates to nothing.

Electronics Design - Lumping in motors/actuators along with the computing hardware since for both of them you'll usually be relying on commercial off-the-shelf actuators and computational hardware, but many, many problems are being stop-gap solved by developing specialized actuators and computers until the general purpose stuff advances in performance to where the system needs it.

In sum - learn controls, and learn it really well. Some version of FPE is good as an introductory text, and FPW is a good advanced text/reference to get started. If you need more than that, hopefully you'll find yourself at a college where you can start asking people about nonlinear controls and more.

I like [this one] (https://www.amazon.com/Feedback-Control-Dynamic-Systems-7th/dp/0133496597).