Artificial Intelligence at the Edge

STMicroelectronics

Artificial intelligence (AI) is a set of hardware and software systems capable of providing computing units with capabilities that, to a human observer, seem to imitate humans’ cognitive abilities.

It uses an assembly of nature-inspired computational methods to approximate complex real-world problems where mathematical or traditional modeling have proven ineffective or inaccurate. Artificial Intelligence uses an approximation of the way the human brain reasons, using inexact and incomplete knowledge to produce actions in an adaptive way, with experience built up over time.

Machine Learning (a subset of AI) addresses a variety of problems which occur in everyday life. They can exploit the data provided by sensors present in our environments, homes, offices, cars, factories, and personal items. A widespread model assumes the raw data from sensors are sent to a powerful central remote intelligence (Cloud), thus requiring significant data bandwidth and computational capabilities. That model would lower responsiveness if you consider the processing of audio, video or image files from 100s millions of end devices.

Switching from a centralized to a distributed intelligence system

AI enables much more efficient end-to-end solutions when the analysis done in the cloud is moved closer to the sensing and actions. This distributed approach significantly reduces both the required bandwidth for data transfer and the processing capabilities of cloud servers, leveraging modern computing capabilities at the edge. It also offers user data sovereignty advantages, as personal source data is pre-analyzed and provided to service providers with a higher level of interpretation.

ST has been actively involved in AI research topics and offers different solutions to push AI to the IoT Edge (the sensor node systems) and even down to the sensor IC itself.

Thanks to ST’s new set of Artificial Intelligence (AI) solutions, you can now map and run pre-trained Artificial Neural Networks (ANN) using the broad STM32 microcontroller portfolio thanks to the STM32Cube.AI extension pack, enabling AI on STM32 Arm® Cortex®-M-based MCUs.

The STM32Cube.AI is:

  • Interoperable with popular deep learning training tools such as TensorFlow Lite, Keras, Caffee, Lasagne, ONNX, etc.
  • Compatible with many IDEs and compilers
  • Sensor and RTOS agnostic
  • Allows multiple Artificial Neural Networks to be run on a single STM32 MCU
  • Full support for ultra-low-power STM32 MCUs
  • Use the power of Deep Learning to enhance signal processing performance and increase productivity in your STM32 application. Create and map Artificial Neural Networks onto your STM32 (optimized code automatically generated) instead of building hand-crafted code.

ST then offers advanced MEMS sensors, such as the IMUs LSM6DSOX, LSM6DSRX and the industrial grade ISM330DHCX.  These devices (and every future MEMS device with X at the end of the part number) contain digital functions optimized to run Machine Learning algorithms that allow sharing data processing between the IMU and the host processor. This approach allows to further reduce the power consumption of the system as for typical tasks the sensor ASIC can consume down to 0.001 times the power of an MCU. Read more about the embedded Finite State Machine (FSM) and Machine Learning Core (MLC) feature in these IMUs and related development tools in these linked Application Notes below.

 

 

Data collection in Industry 4.0 applications is an essential part of the monitoring process and helps ensure factory floor machines run smoothly. Continuous condition monitoring techniques are normally used on equipment such as compressors, pumps, and motors.

Predictive Maintenance (PdM) is based on Condition Monitoring, abnormality detection and classification algorithms, and integrates predictive models which can identify the remaining machine runtime left, according to detected abnormalities. This approach uses a wide range of tools, such as statistical analysis and Machine Learning to monitor the state of the equipment.

Our ultra-low-powerSTM32 Arm® Cortex® M4/M33/M7 based microcontrollers and STM32 Arm® Cortex®-A7® microprocessor series with floating points capabilities can process sensor data at the edge. The STM32Cube.AI toolchain allows to implement Neural Networks and machine learning to implement a deep learning approach.

ST offers high-performance, cost-competitive sensors with a 10-year supply guarantee (longevity program), including accelerometers and ultra-sound analog microphones to enable vibration analysis from simple Pass/Fail monitoring to high-accuracy, frequency-based data analysis, as well as a range of environmental sensors for temperature, humidity and pressure

 

STM32 Portfolio

STM32 Solutions for Artificial Neural Networks based on the STM32 Arm® Cortex® M4/M33/M7 based microcontrollers and STM32 Arm® Cortex®-A7® microprocessor series 

Buy now

 

STM32 Connectivity Solutions

The  STM32 Microcontroller (MCU) portfolio also features wireless connectivity solutions including our ultra-low-power STM32WL System-on-Chips:  STM32WL and dual-core STM32WB microcontroller series. The STM32WL SoC is a multi-protocol and open wireless MCU platform able to run the LoRaWAN® protocol through LoRa® modulation, as well as other ad-hoc protocols based on LoRa®, (G)FSK, (G)MSK or BPSK modulations. The multi-protocol wireless STM32WB MCU platform can run  Bluetooth™ 5.0, OpenThread, ZigBee 3.0 and IEE 802.15.4 communication protocols concurrently.

View presentations:


STM32Cube.AI

Thanks to a new set of Artificial Intelligence (AI) solutions from ST you now have the possibility to map and run pre-trained Artificial Neural Networks (ANN) on the broad STM32 microcontroller portfolio. The STM32Cube.AI is an extension pack of the widely used STM32CubeMX configuration and code generation tool enabling AI on STM32 Arm® Cortex®-M-based microcontrollers. To access it, download and install the STM32CubeMX (version 5.0.1 onwards)

Learn more


ISM330DHCX

The ISM330DHCX is a system-in-package featuring a high-performance 3D digital accelerometer and 3D digital gyroscope tailored for Industry 4.0 applications

Buy now


SensorTile Wireless Industrial Node (STWIN)

The STWIN SensorTile wireless industrial node (STEVAL-STWINKT1) is a development kit and reference design that simplifies prototyping and testing of advanced industrial IoT applications such as condition monitoring and predictive maintenance.

View Parts 


SL-PREDMNT-E2C

Edge processing enabling Condition Monitoring and Predictive Maintenance: quick start for end to end architecture based on wired Smart Sensor Nodes and Gateway.  The Predictive Maintenance Platform (PMP) is a condition monitoring application for the operating conditions of industrial equipment.

View Parts


FP-IND-PREDMNT1

FP-IND-PREDMNT1 is an STM32Cube function pack including dedicated algorithms for advanced time and frequency domain signal processing and analysis of 3D digital accelerometers with flat bandwidth up to 5 kHz.

View parts

 

SPC5Studio.AI extension pack

SPC5-STUDIO-AI is the artificial intelligence (AI) plug-in of the SPC5-STUDIO development environment supporting the SPC5 "Chorus" automotive MCU family. It provides neural network architects a seamless way to generate, execute and validate pre-trained NN models on automotive MCUs.

Learn More

 

Datasheets

 

Other Products:

DSH-PREDMNT

The Predictive Maintenance Dashboard is a cloud application based on AWS services. It provides a highly functional and intuitive interface that is tailored for the collection, visualization and analysis of condition monitoring data from motion and acoustic vibration sensing elements, as well as temperature and other environmental data. You can use the dashboard to plot and graph real-time and historical data, monitor critical operating conditions such as running temperature, and set thresholds for automatic warnings when key parameters exceed acceptable limits.

Learn More

 

 

AI at the Edge - Industrial

Learn More

AI at the Edge - Transportation

Learn more

Hackster Supplier Blog

Learn More

Element14 Supplier Tech Spotlight

Learn More

  

 

stmicro-gridbox-RR-unboxing-video

Display portlet menu

STMicroelectronics SensorTile.box Kit

Sensors allow for complex environmental information to be converted into meaningful data for processing. Learn how the SensorTile.box Kit is advancing edge AI by delivering quality data.

Watch now