Personal review of internship in robotics
Last week I've just completed 6-month internship in robotics research. 2 new things this time were the field (robotics) and working in U.S. I was able to get a very good exposure in robotics in various tasks as I personally hoped, ranging from simulation to navigation on real robots. After working intensely (I already had 1 year part-time work experience in robotics lab, but not this intense), I should say that robotics has fascinated me far more than I expected, and the experience of this internship was enough for me to determine to pursue my 2nd career in robotics.
I'm just listing the thoughts, could be valuable to my future job searching, should not necessarily true. That said, comments from the experienced are the most welcome.
N-1. Researchers are in the greatest need (-> Engineers are not that much?)
Researchers say that it would be long before when robots play the role as in the way they are expected. A lot of them even say it would still take 40 or 50 years by then, similar to how long the cars took to be put into our daily life. So except for very few companies who are already successful commercially (e.g. iRobot), robotics is still at the research & development phase per se, where there seem to be so far no/less need for product development engineers.
Considering the situation, being a software developer w/o really having knowledge & skill in robotics might not help much for working in robotics field. Instead, being able to individually execute research, or at least being able to communicate well using the concept and terms in robotics domain, is necessary, which I have caught up much during the internship but not sufficient yet. So if I choose to stay in robotics, I'd have to either (A) get a PhD, (B) study by myself for the moment, get a job in robotics and accumulate experience, or (C) do B and later A.
N-2. Need to learn robotics tech components in order to become a robot systems designer
Not surprisingly, you need to know the mechanics if you want to design wheels. Learning the domain is necessary no matter which tactics I choose from 3 choices I mentioned earlier.
The followings are the tech components that someone advised me this time. These are basic fields, but still not exhaustively depicting the entire domain.
- Math (linear, probabilistic, control theory)
- Realtime programming method (I interpret as embedded multi-thread programming)
Also I felt in need:
- Programming language (Matlab (to communicate with other researchers) and C++ especially boost library that I had a trouble this time)
N-3. Figure out the standpoint in robotics engineering
Relative to N-2, I'm curious how skillful and competitive I am in robotics industry/academia, but apparently there's no clear measure out there. I didn't receive an explicit evaluation at this internship either. So I'll always have to figure out somehow by myself at least while I'm not employed.
Two thing for sure is I don't have experience in hardware, and also integrating a robot by myself. Someone in the industry pointed out that esp. young engineers need experience of making a robot by themselves so that they can have a global view of systems. Otherwise those engineers tend to fall back only on local software component without considering the system itself and taking hardware constraints into the design. I totally agree with it.
N-4. Do not forget the enthusiasm both in robotics and in software development I found this time.
So the reason I got fascinated to robotics this much is, simply put, physical interaction b/w systems and humans such as integrating sensors, mechanical arms etc. is fun! And it's been 4 years since I quit the last job where I was intensely involved in software system integration, and I felt again that building systems (regardless pure software or hardware related) is what makes me motivated the most. I've never felt tired by the job these 6 months and I always wanted to work more. I think I should elaborate this more for myself soon. Anyway, if I was still holding a certain job position without quitting it and becoming a grad student, then I could have not thought freely about shifting toward the different tech area in my 2nd career like I'm actually doing now.
(Optional) Internship might be a bit hard for those who already have professional background...
You're nothing but an intern no matter what level of experience you have. That means even colleagues who're fresh out of undergrad sometimes see you the same/lower level than themselves. Although it wasn't that hard for me that I just had to understand, the issue was they don't often pay attention to my suggestion (I guess it would have been different if I wasn't working as an intern).
I'm just listing the thoughts, could be valuable to my future job searching, should not necessarily true. That said, comments from the experienced are the most welcome.
N-1. Researchers are in the greatest need (-> Engineers are not that much?)
Researchers say that it would be long before when robots play the role as in the way they are expected. A lot of them even say it would still take 40 or 50 years by then, similar to how long the cars took to be put into our daily life. So except for very few companies who are already successful commercially (e.g. iRobot), robotics is still at the research & development phase per se, where there seem to be so far no/less need for product development engineers.
Considering the situation, being a software developer w/o really having knowledge & skill in robotics might not help much for working in robotics field. Instead, being able to individually execute research, or at least being able to communicate well using the concept and terms in robotics domain, is necessary, which I have caught up much during the internship but not sufficient yet. So if I choose to stay in robotics, I'd have to either (A) get a PhD, (B) study by myself for the moment, get a job in robotics and accumulate experience, or (C) do B and later A.
N-2. Need to learn robotics tech components in order to become a robot systems designer
Not surprisingly, you need to know the mechanics if you want to design wheels. Learning the domain is necessary no matter which tactics I choose from 3 choices I mentioned earlier.
The followings are the tech components that someone advised me this time. These are basic fields, but still not exhaustively depicting the entire domain.
- Math (linear, probabilistic, control theory)
- Realtime programming method (I interpret as embedded multi-thread programming)
Also I felt in need:
- Programming language (Matlab (to communicate with other researchers) and C++ especially boost library that I had a trouble this time)
N-3. Figure out the standpoint in robotics engineering
Relative to N-2, I'm curious how skillful and competitive I am in robotics industry/academia, but apparently there's no clear measure out there. I didn't receive an explicit evaluation at this internship either. So I'll always have to figure out somehow by myself at least while I'm not employed.
Two thing for sure is I don't have experience in hardware, and also integrating a robot by myself. Someone in the industry pointed out that esp. young engineers need experience of making a robot by themselves so that they can have a global view of systems. Otherwise those engineers tend to fall back only on local software component without considering the system itself and taking hardware constraints into the design. I totally agree with it.
N-4. Do not forget the enthusiasm both in robotics and in software development I found this time.
So the reason I got fascinated to robotics this much is, simply put, physical interaction b/w systems and humans such as integrating sensors, mechanical arms etc. is fun! And it's been 4 years since I quit the last job where I was intensely involved in software system integration, and I felt again that building systems (regardless pure software or hardware related) is what makes me motivated the most. I've never felt tired by the job these 6 months and I always wanted to work more. I think I should elaborate this more for myself soon. Anyway, if I was still holding a certain job position without quitting it and becoming a grad student, then I could have not thought freely about shifting toward the different tech area in my 2nd career like I'm actually doing now.
(Optional) Internship might be a bit hard for those who already have professional background...
You're nothing but an intern no matter what level of experience you have. That means even colleagues who're fresh out of undergrad sometimes see you the same/lower level than themselves. Although it wasn't that hard for me that I just had to understand, the issue was they don't often pay attention to my suggestion (I guess it would have been different if I wasn't working as an intern).
 |
| To debug robot system is a complicated task; you have to care various components - hardware to software. I used ROS in the workplace - open source robotics platform that's becoming a standard for robotics research. This image shows an example plot of all running processes in a single system that I worked on. |
Comments