Robots, rockets, and more! Control theory in 10 minutes!

on YouTube ↗10m

" summary

Live coding builds a simulated elevator and tries controllers that actually run: bang-bang control flips a force back and forth and oscillates, proportional plus derivative smooths motion but stalls, and explicitly apply 980 Newtons to cancel gravity for a 100 kg car yields perfect tracking. System uses state-space matrices A and B and linear algebra tools (eigenvalues, MATLAB-style computation) to predict stability and design an optimal linear controller.

AI-written summary · about

" chapters (4)

1. 0:00Intro
2. 0:20What is control theory
3. 1:48Bangbang control
4. 5:40Statespace representation

" related talks

• 30mDeveloping Control for Mobile RobotsYan Tsybulkyn · Highload fwdays'17
• 17mBeyond Automation: AI and the Evolution of Robotic IntelligenceMarvin Wiedemann · AI & Data Summit and Quantum Summit 2025
• 23mAI.js: Robots with Brains!Raquel Vélez · JSConf US 2013
• 10mPyTorch in RoboticsYisong Yue · PyTorch Developer Conference 2019
• 50mOptimal control, no! Than what? (...)Tomasz sikora · JDD 2018
• 41mCompetitive Algorithms for Online ControlYisong Yue · Mathematics of Online Decision Making 2020
• 49mUsing Deep Learning for Perception in Autonomous Systems: A Perspective from Control TheoryClaire Tomlin · Frontiers of Deep Learning 2019
• 30mBuilding, designing, teaching and training simulation environments for Machine LearningParis Buttfield-Addison · PyCon AU 2019