Building Products Using Edge AI - TinyML on MCUs

Explore the intricacies of building products using Edge AI and TinyML on microcontrollers in this 33-minute tinyML Talks webcast featuring Stuart Feffer from Reality AI. Discover how to leverage AI for sensor selection, placement optimization, and component specification determination while minimizing data collection costs. Learn to generate sophisticated, explainable machine learning models based on sensor data automatically using Reality AI Tools 4.0. Gain insights into the differences between project execution and product development through real-world case studies, including condition monitoring for an industrial blower. Delve into topics such as AI Explore results, feasibility proof, BOM optimization, production ML model building, and data readiness considerations throughout the product development lifecycle.

Intro
Building products vs doing projects
Condition Monitoring for an Industrial Blower
Al ExploreTM results - Explainability
Phase 1: Prove feasibility
Best sensors? Best location?
BOM Optimization - Channel Selection
BOM Optimization - Sensitivity
Phase 2: Design the product
Build production ML model
Data Readiness - Consistency
Data Readiness - Quality
Data Readiness - Category Coverage
Data Readiness - Time Coverage
Product development lifecycle