Roadmap to Robotics — Part 3
--
Read the first part of Roadmap to Robotics Series here — https://kajal-gada.medium.com/roadmap-to-robotics-part-1-10c7274c1526
Level 3 Expert
In this section, I am listing down topics, algorithms, libraries, etc specific to sub-fields (including the ones from Level 2). This is not an exhaustive list. Feel free to send me a message on Linkedin to add further.
Mindset & Pro Tips
- You first learn the basics of multiple sub-fields (Level 2) and then choose a sub-field (or combination) to learn in depth.
- Re-Implement a well-known paper or well documented project online, which once implemented, can be modified around and improved on. Best way to internalize the learnings, understand its pros and cons.
- Make use of an existing library into your project. Note use of functions and their limitations.
Math
When it comes to Robotics & Math, there are areas where you don’t need math and there are areas where you need it. And then there are areas of Robotics which depend on Maths but you can get away with it by using libraries.
If you want to be an expert in Robotics, it is a good idea to level up your Mathematics knowledge:
- Multivariable Calculus
- Probability
- Linear algebra
Quick Note on Navigation
Navigation has everything to do with how the robot will get to its destination. So inherently the robot has to know the map (or build one on the go), figure out where it is, create a path to get where it wants to go and deal with any changes — aka now there is a wall in front of me which was supposed to be free space.
You can choose to be an expert in one of the subfields — mapping, localization, path planning, dynamic obstacle avoidance, etc or even a combination like SLAM.
But before you do that, it is a smart idea to have the basic knowledge of how all the elements of navigation work. Because it is an interdependent system.
Path Planning
- Map Representation
- Breadth first search (BFS) & Depth first search (DFS)
- Dijkstra
- Global & local planning
- Cell Decomposition based Approaches
