The Future of Analog Neural Networks and Integration with Control Systems

Hello everyone,

I am a second-year Control and Automation Engineering student. Since my childhood, I’ve been deeply involved in the "maker" community, working with microcontrollers and basic electronics. Currently, I am working on Deep Learning projects, but my primary interests remain rooted in Control Theory and Analog Electronics.

Recently, I’ve been researching Analog Neural Networks (ANNs). I am currently teaching myself how to design a neuron’s logic using analog circuitry (op-amps, resistors, etc.). However, I have some questions regarding the career path and the industry's direction in this field:

1. Relevance: Is Analog Neural Network research considered a "legacy" topic, or is it gaining modern traction? With the rise of Neuromorphic Computing and In-Memory Computing, does focusing on analog implementations of AI provide a viable career path?
2. Control Theory Integration: Can I bridge the gap between Deep Learning and Control Theory through analog hardware?
3. Feasibility: For someone at my level, is diving deep into analog AI hardware a sound investment of time, or should I stick to the digital/software side of ML?

I would love to hear from anyone with experience in VLSI, Neuromorphic systems, or advanced Control Systems. As a sophomore, I am eager to align my self-study with the future of the industry. Additionally, as a sophomore student, I am open to any career or project advice. What specific sub-topics or emerging technologies should I focus on to stay ahead in the next 5-10 years?

Thank you in advance for your guidance!