Industry 4.0: Adding Intelligence to Automation with SensiML's Analytics Toolkit

by Chris Rogers, SensiML - 11-18-2021

Since the beginning of the industrial revolution, increasingly sophisticated levels of automation have driven a massive increase in the global standard of living. The next wave, often called "Industry 4.0", will further enhance efficiency and productivity for a wide range of automation applications.

onsemi designed the RSL10 Sensor Development Kit to help drive the transition to Industry 4.0. It features a small-yet-comprehensive platform that enables developers to quickly build connected edge-node applications using leading-edge sensor technology and the industry's lowest power Bluetooth^® Low Energy radio - the RSL10. The RSL10 radio SoC includes an integrated 32-bit Arm® Cortex®-M3 processor, a 32-bit dual-Harvard DSP core, multi-protocol wireless communications support, integrated power management, flash, RAM, and a host of peripherals and interfaces. The kit features a range of easily configurable sensors, including environmental, inertial, geomagnetic, and other sensor types.

RSL10 Sensor Development Kit

onsemi and SensiML recently announced a complete AI/machine learning solution, complementing the extremely capable RSL10 Sensor Development platform with SensiML's Analytics Toolkit. Developers using our combined solution can easily add low latency local AI-based algorithms to their manufacturing, process control, robotics, or predictive maintenance applications without needing to have expertise in data science or AI. The Analytics Toolkit takes system inputs and auto-generates code to create intelligent sensing embedded endpoints that process data locally to provide critical insights quickly and take appropriate action in real-time.

Having these smart local IoT endpoints allows companies to monitor their process flows and optimize them continuously or to take other actions such as machine maintenance preemptively to save downtime and its associated high costs. At the same time, having local intelligence can remove unnecessary work by avoiding maintenance that isn't necessary or beneficial.

Some companies have chosen to implement cloud-based AI, but, in most cases, having the AI processing located at the endpoint is a more desirable solution. It converts a high volume of raw sensor data from tens to thousands of sensing points and from a wide range of sensor types such as audio, motion, acceleration/vibration, current, or voltage into critical, actionable insights quickly without consuming a lot of power and without bogging down the network. It also avoids the latency, security, and reliability issues involved with transmitting information broadly to the cloud.

The onsemi RSL10 Sensor Development Kit, together with the SensiML tools, lets developers test their ideas with an easy-to-use but powerful system well suited to implementing edge AI IoT applications. The RSL10 is highly integrated but consumes little power, and the carefully designed SensiML software provides high-quality AI analysis on local processors without adding significant power consumption. The combined result is a compact but sophisticated platform that's perfectly suited for driving next-generation Industry 4.0 solutions.

Learn more about edge AI applications by registering for our upcoming technology webinar,How to Create Edge AI Models for Industrial Sensor Nodes.