> eventually program something that to some extent can think for itself. Depending on how realistic you are as to what you are trying to do, this could be easy

> eventually program something that to some extent can think for itself. Depending on how realistic you are as to what you are trying to do, this could be easy or enormously difficult. Artificial Intelligence in the sense of what exists in Movies is either hard/close to impossible, despite several breakthroughs due to large scale processing. However, there is a significant body of work that has been done in this field that may be off interest. 1. How does a Quadrotor fly? This could be an easy problem of getting an off the shelf quadrotor? Alternatively, you could learn a bit of control theory, get some modeling work done and figure out how to build your own Quadrotor? This is typically hard and you would need some math in your belt (Linear Algebra, Calculus, Dynamics and Control Systems ). Something off the level of a senior in an EE program. 2. Now that you have got the thing flying, what do you need next? You want it to follow someone. Let us make the problem simpler. Say, you want it to follow a line on a path. This could be as easy as building a program that tells the Quadrotor to keep following the line, which you will detect using a camera coupled with some edge detecting software. 3. How do you go after this? How do you make sure it follows someone? Same principle, instead of detecting edges, you have to make sure it detects a small moving object. Once you have figured that out, you iteratively progress till you figure out how to make it follow a person. This is not the same as traditional Machine Learning. Although the same principles, a math heavy background with good programming skills can help you translate and make the switch. If you are into Robotics or ML, a heavy math program coupled with good coding skills will always help you. It doesn't have to be an EECS program. I did EECS. I know some excellent people who came from a Mechanical background. As far as courses are concerned, just go to college and keep working on your own personal projects, the structure of any Engineering program should be sufficient for your purpose.