Sensor Integration and Battery Life Lead AI-IoT-Embedded Confluence

April 25, 2024

Story

One of the key messages to come out of embedded world in Nuremburg, Germany this year was that Embedded computing is becoming increasingly more entwined with IoT systems and AI applications and that customers are looking for these all-in-one solutions at every opportunity. And it's no surprise that the first fertile furrows to plant are the sensor integration and energy efficiency rows. 

In recognition of that, and often leading the charge, big companies are announcing new holistic solutions that add AI and IoT capabilities to embedded products or drive embedded technologies to support critical IoT and AI functions and needs.

Let’s take a look at a few recent examples.

Sensors

POLYN Technology, a fabless semiconductor company specialized in Neuromorphic Analog Signal Processing (NASP), just announced at Hannover Messe its vibration monitoring solution, based on its VibroSense for Machine Health Monitoring. The company says it offers a reduction in vibration signal data by more than 1000 times. According to the release, the solution implements vibration monitoring in an analog neuromorphic design that can identify a machine’s distinctive vibration waveform, which a neural network can distinguish from other vibrations and extract from any background noise. What’s more, all this processing reportedly happens with ultra-low power consumption at the edge thanks to the VibroSense chip.

STMicroelectronics is growing its edge-AI and IoT sensor family, according to a recent announcement, by adding an inertial module for intensive movement analysis. The LSM6DSV32X 6-axis inertial module (IMU) reportedly has as accelerometer full-scale range of 32g and a 4000 degrees-per-second (dps) gyroscope that together can measure both movement and impact. The company says that this module is ready for even generations of edge-AI applications and can be used in consumer wearables, asset trackers, and impact and fall alarms.

Interestingly, even with its powerful embedded, IoT and AI features, the LSM6DSV32X reportedly is designed to be efficient and conservative with battery power, thanks to ST’s Sensor Fusion Low-Power (SFLP) algorithm and adaptive self-configuration (ASC) that autonomously reconfigures sensor settings in real-time to optimize performance and power.

And speaking of power conservation, let’s discuss….

Energy

Silicon Labs has released a new efficient energy harvesting solution to allow battery-free IoT, according to a recent announcement. The new xG22E family of wireless SoCs is the company’s first series designed to operate within the ultra-low power envelope required for battery-free, energy harvesting IoT applications. The new family includes the BG22E, MG22E, and FG22E SoCs, all of which will enable IoT devices to use Bluetooth Low Energy, 802.15.4-based, or Sub-Ghz wireless connectivity for both battery-optimized and battery-free implementations. Further, they enable environmental energy harvesting from ambient light, radio waves, and kinetic motion.

What’s more, the company has partnered with Power Managed Integrated Circuits (PMIC) company e-peas to develop two energy harvesting shields for the new xG22E Explorer Kit from Silicon Labs. With this kit, developers reportedly can customize peripherals and debugging options to best match any given application and implement the energy harvesting shields.

This kind of intersectional development, bringing embedded, IoT, and AI solutions together into single products and systems is the future of the horizontal technology enablement layer, as every vertical industry realizes that analysis, processing, efficiency, and connectivity are now critical infrastructure. It's going to be exciting to see what the harvest is like, and what gets planted next.