This reply will focus mainly on the academics.
Since you've mentioned exoskeletons, knowledge of kinematics and dynamics is imperative.
Rotation Matrices, Forward/Inverse Kinematics, Denavit - Hartenberg Parameters, Lagrangian Mechanics are a few fundamental concepts one should be familiar with. Their applications mostly pertain to robotic manipulators (arms), which are what members of exoskeleton's are modeled after.
They're covered extensively in the classic in the field textbook
Robotics Modelling, Planning and Control by Siciliano, Sciavicco, Villani, Oriolo
https://link.springer.com/book/10.1007/978-1-84628-642-1They also require some prior knowledge of linear algebra to safely navigate through, so make sure you've achieved at least some math literacy before diving into them.
Speaking of navigation, if you're interested in motion planning i.e. how to optimally (safely and efficiently) go from point A to point B, what you read is
Planning Algorithms by Steven M. Lavalle
http://lavalle.pl/planning/for various ways the math people have came up with to solve this. Many cool applications in fields outside of robotics like in Computer Graphics/Animation too.
And btw, if there's one paper you'll absolutely have to read if you find yourself more interested in motion planning is
Sampling-based Algorithms for Optimal Motion Planning by Sertac Karaman and Emilio Frazzoli
https://arxiv.org/pdf/1105.1186.pdfin which the authors have revised two very popular path planning algorithms by making them significantly more optimal than their initial implementations were, and are part of many decision making systems that are involved in any type of mechanical movements.
Some other comments talked about more advanced disciplines in the field like State Estimation or Reinforcement Learning but I believe the aforementioned (kimenatics/dynamics/motion planning) are the bare minimum before diving into even more advanced math-heavy concepts.