Articles

This document serves as a post-webinar guide for following along with the hands-on workshop on how to create an acoustic recognition algorithm using the SensiML Toolkit and Silicon Labs' MG24 silicon with MVP.

This comprehensive overview defines the smart edge AI approach for building IoT sensing applications and explains the many benefits of creating intelligent endpoints. It then discusses automated machine learning workflows (known as “AutoML”) and reviews the key stages of the development process from modeling to prototyping. Most importantly, the paper also gives developers new to AI and machine learning practical guidance and examples for implementing designs using AutoML tools.

"Internet of Things (IoT) endpoints have, until very recently, been focused on acquiring raw sensor data and feeding the associated data back towards the “core” of the network for processing. The network’s core data processing resources, whether local or cloud connected, have traditionally been the focus for analyzing the data to identify meaningful events and making decisions based upon the results. As Artificial Intelligence (AI) has become more prominent for sensor data analysis, these same centralized processing resources would seem to be the obvious place for this processing to be implemented."

This document describes the protocols, supporting documentation, and reference code needed for SensiML users to adapt their customized board firmware and SensiML software configuration to properly interface and work in a manner consistent with the supported evaluation kits.

"While much has been made of AI and ML analytics in the cloud and in edge computing, the overlooked opportunity for analytics contributions at the extreme edge (i.e. the sensor processor itself) remains largely untapped. This presentation - delivered at Sensors Expo 2018 in San Jose, CA - delves into the current and forthcoming advancements in extreme edge processing as part of an overall IoT network solution and the benefits and challenges of a more active role for sensor analytics processing."

"Counter intuitively, it is often more efficient to simply leave a sensor on permanently, waiting to identify useful information, e.g. an accelerometer in a step counter application. Our sensor system must intelligently determine which data is worth transferring to the cloud and thereby efficiently utilize the available bandwidth and power. The key is for local on-sensor processing to discard most of this superfluous data autonomously and thus save valuable system driver capacities."

This research paper from SensiML's early pre-spinout activities in edge ML and gesture recognition demonstrates the methodologies developed and since evolved by SensiML for optimized classification of gesture command and control applications.