Machine Learning Foundations: Prototyping on the Edge
1h 22mBeginner2023-05-11
Authors

Robert Gallup
Technologist, Designer, Maker, and Speaker
Course details
As AI and compute power continue to grow, so does the need and opportunity for decentralized intelligence or intelligence “on the edge.” Prototyping allows product designers, product managers, and developers to focus on exploring product ideas, without a deep knowledge of theory, and tools and hardware now available make this possible. In this course, technologist, designer, and speaker Robert Gallup guides you through the process of prototyping machine learning on the edge. Robert begins with an overview of machine learning, including machine learning elements, workflow, and ethics. Next, he then discusses prototyping, and presents a scenario and plan for the voice recognition prototype developed in the course. Robert then walks you through how to prepare software and hardware, as well as how to define, train, and deploy a machine learning model. Finally, he shows you how to extend your prototype and discusses considerations in making the move from prototype to production.
Skills covered
ArduinoPrototypingMachine LearningProduct and ManufacturingArtificial Intelligence (AI)One-Off
Concepts
0. Introduction
- 01 - Welcome
- 02 - Course resources
1. Machine Learning Overview
- 03 - Introduction to machine learning on the edge
- 04 - Machine learning elements
- 05 - Machine learning workflow
- 06 - Machine learning ethics
2. Prototyping
- 07 - Prototyping overview
- 08 - The prototype plan
3. Preparing Software and Hardware
- 09 - Set up the Arduino IDE
- 10 - Arduino hardware
- 11 - Run a machine leaning demo
4. Defining and Training an ML Model
- 12 - Create a project in Edge Impulse
- 13 - Acquiring data
- 14 - Define the machine learning pipeline
- 15 - Training the model
5. Deploying the ML Model
- 16 - Deploying to computer
- 17 - Deploying to Arduino
- 18 - Running the model
6. Extending Your Prototype
- 19 - Adding LED functionality
- 20 - Adding keyboard functionality
- 21 - Challenge - Add another keyword
- 22 - Solution - Add another keyword
Conclusion
- 23 - From prototype to production
- 24 - Final thoughts