Justin Grammens

Camp Counselor

TinyML : Let’s Put Some Intelligence Into IoT!

Event Logo

Monday, July 25, 2022 - 8:15 PM UTC, for 2 hours.

AT THAT (In-Person Only) Workshop (pre-conference)

Room: D

TinyML

In this session we will review what TinyML is, the advantages it has over other more traditional approaches, discuss some of the many applications where it fits and where it might not, how to get started by walking through a few real-world working examples that I have created. We will show the entire lifecycle from data ingestion using sensors, data cleaning, model creation/training, deployment, and then finally some predictions on the device. We will be using sensor data and executing code on device. If done in person, I’m thinking we would run this as a hands-on workshop with attendees either: 1. Purchasing hardware before the event 2. Using their cell phone as the device and we’ll use the sensors on the device with a mobile application. In either case, we would wire together a few quick prototypes and use much of the work done by a company called Edge Impulse to help us with the heavy lifting. I’m open to a variety of options depending on what you think the audience would like. I just really dig this stuff and would like engineers to see how powerful TinyML is and how accessible it is today to people of all ages.

Agenda

If done in person, I’m thinking we would run this as a hands-on workshop with attendees either: 1. Purchasing hardware before the event 2. Using their cell phone as the device and we’ll use the sensors on the device with a mobile application. In either case, we would wire together a few quick prototypes and use much of the work done by a company called Edge Impulse to help us with the heavy lifting. I’m open to a variety of options depending on what you think the audience would like. I just really dig this stuff and would like engineers to see how powerful TinyML is and how accessible it is today to people of all ages.

Prerequisites

Laptop, Cell Phone, Experience with C/C++ or Python

Take Aways

  • Capturing Data from the Physical World
  • Training a Machine Learning Model
  • Intelligence on IoT Devices

Supporting Resources

favorited by:
Bryan Shannon