#controlsystems #roboticsengineering #drone #nonlinear #datascience
Robotics engineering is composed of control systems with nonlinear dynamics. Like data science, modeling the data from the system is the game rule to optimize the robot’s controller.
———————————————————————————–
📝In-video sources
➡️ UTIAS paper:
➡️ SiNDY paper:
➡️ Steve Brunton:
———————————————————————————–
🌏 Find Me Here:
🔥Linkedin:
🔥Discord:
🔥ubicoders:
🔥Blog:
🔥GitHub:
———————————————————————————–
⏲️Time Stamps:
0:00 open
2:30 simple example with free fall motion
3:43 why is system identification required?
4:50 the system identification is about finding a model from the data.
5:06 how about nonlinear system?
5:27 sindy, sparse identification of nonlinear dynamical systems
6:34 robotics research vs development
date 2024-12-15 01:08:06
views 7195
author UC2RxqAYQt-LBs3paWv78rLA
source
What’s the Difference?: Robotics Research vs. Real-World Development
As a robotics engineer, I’ll dive into the difference between research and development in the field. In this video, I share my personal experience with tuning a controller for a tail seater UAV, which took me a couple of months, whereas a similar task in software development would only take a few minutes. The key to this is system identification, which involves finding the right numbers to describe the system’s behavior.
In this case, I had to figure out the relationship between the motor’s PWM input and its proportionate force output. To do this, I set up a basic lab with a strain gauge and 3D-printed motor mount to measure the proportion force while varying the PWM input. The goal is to find the inverse function of this relationship to generate the correct PWM input from the main controller.
The video also touches on the difference between research and development, with research focusing on discovering new algorithms for controlling new robots, while development takes those research papers and builds robots and sells them. I also share my take on the importance of building upon existing research papers, like Cindy, which helps identify the governing equations of a system from time-series data, making it a powerful tool for researchers and engineers.
To decide between research and development, ask yourself which one you enjoy more: pushing the boundaries of what’s possible and solving theoretical problems, or applying existing tools to build practical systems that make an impact. Both are essential and interconnected, and it’s an exciting time to be in robotics.
