Introduction To Ai/Ml Motion Control

Focusing on RL Autonomous Systems

What you'll learn

List the benefits of AI/ML automation products, in comparison with classical PID automation products

Identify the differences between new AI automation product and common PID motion-controlled product

Understand the required team and the development process of AI/ML automation products

Plan objectives, action items, resources and cost of a simple startup RL motion-controlled product

Requirements

Interest in next generation development process of autonomous AI/ML mechatronics products

Experience in automation, such as, education, engineering, marketing, design, manufacturing and application

Description

This course extends a Udemy course with over 10k students, "Introduction to Mechatronics, including AI/ML features", which summarizes a Graduate course taught by the author at Stony Brook University.  This course introduces an exciting topic of AI/ML motion control, based on SAAR Inc. few years of activity in AI/ML mechatronics. This course is intended for engineers, scientists, marketing, business and investors with interest in autonomous product development for virtually any field. It explains the basic principles of commonly used Supervised Learning (SL) Labeled Datasets. It highlights the more difficult topic of Reinforcement Learning (RL) as used in autonomous motion control systems.  It explains how RL maximizes a Reward function based on Static, Kinematic, Dynamic, and Motion Control simulation, with sensed State of force, position, velocity, sound and image sensors, all as an input to the Neural Network (NN) Controller. The course illustrates how the NN is mathematically being trained, and after being trained, how the output Actions of the motion control drive the system actuators in an optimal Rewarded way. The course highlights the similarity to an old PID motion control, which dominates existing automation products. The course describes the technology pyramid of Mechanical, Electrical, Software, System Analysis, AI/ML, Data Storage and IoT communication, jointly with market, business and investment needs, which move new autonomous systems towards Technological Singularity.  Finally, the course highlights key companies which sell RL products, and those who provide AI/ML training tools and data storage clouds, for global IoT deployment of trained NN controllers into automation products all over the world.     

Overview

Section 1: Introduction

Lecture 1 Motivation

Lecture 2 Who is using RL motion control

Section 2: Motion Control

Lecture 3 The AI pyramid of Values

Lecture 4 Types of AI motion Control

Lecture 5 PID motion Control

Lecture 6 RL System Blocks

Lecture 7 The difference between PID and RL

Section 3: Development Tools

Lecture 8 System Analysis

Lecture 9 Reward Function

Lecture 10 RL Agents, Critics, Actors

Lecture 11 NN Optimization, Belman Equation

Section 4: Conclusion

Lecture 12 Deployment of NN

Lecture 13 RL Startup Options

Mechatronics engineers ME, EE, CS, System Analysts, who are excited to start designing AI/ML motion control,Business, Marketing and Investment managers in an automation business, who value AI/ML in their market,Industrial, healthcare, agriculture, semiconductor, electronic equip' manufacturers who consider AI in their tools,Professors, teachers and students in high schools, colleges and universities who plan AI/ML courses and labs