I started my undergraduate degree in Electrical and Electronics engineering with plans of majoring in the same. But a couple of semesters into engineering, I felt like not restricting myself to one stream. I wanted to learn mechanical engineering, computer science, and machine learning, but that was not feasible within a classic curriculum. So, I started exploring co-curricular options and joined the Robotics club in my college. The club comprised a handful of undergrads from different disciplines of engineering learning and working together to build robots and participating in various competitions with their robots. I enjoyed working with that multidisciplinary team and spent most of my after-class hours working in the Robotics club on different types of robotics ranging from a simple line following four-wheeled robot to an autonomous guided vehicle to carry carts on an industry floor. That experience and skills led me to a robotics internship in my junior year in a field robotics lab. After thoroughly enjoying my robotics internship experience, I ended up going back to the same lab after my undergraduate degree and getting my master’s degree in robotics.
Advice to students interested in this career path
I would advise students interested in robotic software to start by getting their hands dirty with robotics code and not wait for the perfect opportunity. There are plenty of open-source resources and tutorials out there. There is no single major that can land you a robotic software job. Robotics software engineers can come from various backgrounds. I have co-workers with mechanical engineering degrees, computer science degrees as well as electronics degrees. I believe the key to being an excellent robotics engineer is to be curious about how different things work and to have the ability to collaborate with people from different technical backgrounds.
Students interested in becoming robotics software engineers can start writing simple code to make robots navigate from point A to point B in simulation. Or they can take real-world data from a robot available on open-sourced databases and start programming a solution to identify an object if they are interested in perception. They can make the robot learn to manipulate objects if they are interested in manipulation and so on. Once they get comfortable in this space, they can slowly start testing their solutions on real hardware such as a web camera on an RC car.
