Quick Start Guide

STM32Cube function pack for STEVAL-

STWINKT1B evaluation kit for predictive
maintenance application based on artificial
intelligence (AI)

(FP-AI-PREDMNT2)

Version 1.0 (Oct 15, 2021)

 Agenda

1 Hardware and Software overview

2 Setup & Demo Examples

3 Documents & Related Resources

4 STM32 Open Development Environment: Overview

2
1- Hardware and Software overview
 STWIN development kit - STEVAL-STWINKT1B
Hardware Overview
STWIN - SensorTile Wireless Industrial Node
The STWIN (STEVAL-STWINKT1B) is a development kit and reference design that simplifies prototyping and testing
of advanced industrial IoT applications such as condition monitoring and predictive maintenance. The kit supports
BLE wireless connectivity through an on-board module and Wi-Fi connectivity through a special plugin expansion
board (STEVAL-STWINWFV1), wired RS485 and USB OTG connectivity.

Key Features
• Multi-sensing wireless platform implementing vibration monitoring and ultrasound detection
• Updated version of STEVAL-STWINKT1, now including STSAFE-A110 populated, BlueNRG-M2S module and
IMP23ABSU MEMS microphone
• Ultra-low-power ARM Cortex-M4 MCU at 120 MHz with FPU, 2048 kbytes Flash memory (STM32L4R9)
• Micro SD Card slot for standalone data logging applications
• Option to implement Authentication and Brand protection secure solution with STSAFE-A110
• Wide range of industrial IoT sensors: ultra-wide bandwidth (up to 6 kHz), low-noise, 3-axis digital vibration sensor
(IIS3DWB), 3D accelerometer + 3D Gyro iNEMO inertial measurement unit (ISM330DHCX) with machine learning core,
ultra-low-power high performance MEMS motion sensor (IIS2DH), ultra-low-power 3-axis magnetometer (IIS2MDC),
digital absolute pressure sensor (LPS22HH), relative humidity and temperature sensor (HTS221), low-voltage digital local
temperature sensor (STTS751), industrial grade digital MEMS microphone (IMP34DT05), analog MEMS microphone with
frequency response up to 80 kHz (IMP23ABSU)
Latest info available at
• Modular architecture, expandable via on-board connectors: STMOD+ and 40-pin flex general purpose expansions, 12-pin www.st.com/stwin
male plug for connectivity expansions, 12-pin female plug for sensing expansions
• Other kit components: Li-Po battery 480 mAh, STLINK-V3MINI debugger with programming cable, Plastic box
 STWIN development kit - STEVAL-STWINWFV1
Hardware Overview
Wi-Fi expansion for the SensorTile Wireless Industrial Node (STWIN) kit
The STEVAL-STWINWFV1 expansion board adds 2.4 GHz Wi-Fi connectivity to the STEVAL-STWINKT1B.
Application examples that take advantage of this expansion board can be found also in STSW-STWINKT01 firmware
package as well as in the FP-IND-PREDMNT1 and FP-AI-PREDMNT2 Predictive Maintenance Function Pack.

Key Features
• Wi-Fi adapter for STEVAL-STWINKT1B
• Plugs into STWIN core system board through dedicated 12-pin connector
• Single 3.3 V power supply input
• ISM43362-M3G-L44-E Wi-Fi module:
• 802.11 b/g/n Compatible based on Broadcom MAC/Baseband/Radio device
• Fully contained TCP/IP stack
• Host interface: SPI up to 25 MHz
• Network features ICMP ( Ping), ARP, DHCP,TCP, UDP
• Low power operation (3.3 V supply) with built-in low power modes
• Secure Wi-Fi authentication WEP-128, WPA-PSK (TKIP), WPA2-PSK
• CE, FCC and IC certified
• RoHS and China RoHS compliant
• WEEE compliant
Latest info available at www.st.com
STEVAL-STWINWFV1
FP-AI-PREDMNT2 Software Description FP-AI-PREDMNT2
FP-AI-PREDMNT2 is an STM32Cube function pack that programs the STWIN as an
IoT Edge node, connected to the cloud, able to acquire sensor data, process them and Software Overview
send the results to the DSH-PREDMNT cloud dashboard. It includes dedicated
algorithms for advanced time and frequency domain signal processing and analysis of
3D digital accelerometers with flat bandwidth up to 6 kHz. Overall Software Architecture
Key features
• Compatible with NanoEdge™ AI Studio solution from Cartesiam, to enable AI-
based solution (NanoEdge™ AI library generation is out of the scope of this
function pack and must be generated using NanoEdge™ AI Studio)
• Generic FFT library middleware to enable frequency domain analysis for any kind
of sensor through Fast Fourier Transform (with programmable size, overlapping
and windowing)
• Motion TD library middleware for vibration analysis in time domain (speed RMS
and acceleration peak)
• Configurable alarm and warning thresholds for key parameters
• Compatible with STBLESensor application for Android/iOS, to perform Wi-Fi
configuration and secure certificate provisioning
• Compatible with DSH-PREDMNT web-based predictive maintenance dashboard
for monitoring sensor data and device status
• Easy portability across different MCU families, thanks to STM32Cube
• Firmware modular example based on eLooM (embedded Light object-oriented
fraMework for STM32) to enable code re-usability at application level
• Free, user-friendly license terms

Latest info available at www.st.com

FP-AI-PREDMNT2 6
2- Setup & Demo Examples
 Setup & Demo Examples
HW prerequisites

• 1x STEVAL-STWINKT1B development board

• 1x STEVAL-STWINWFV1
STEVAL-STWINWFV1
• Laptop/PC with Windows 7, 8 or 10
• 2x microUSB cables
• 1 smartphone with ST BLESensor App STEVAL-STWINKT1B
• Wi-Fi Router or access to a Wi-Fi network

MicroUSB Cable

ST BLESensor App
8
 Setup & Demo Examples
HW prerequisites

• How to re-program/debug the STWIN

STLINK-V3MINI

Assembled Kit

Programming cable

9
 Setup & Demo Examples
Software and Other prerequisites

• STM32CubeProgrammer Software
• Download and install STM32CubeProgrammer

• FP-AI-PREDMNT2
• Copy the .zip file content into a folder on your PC. The package will contain source code example (Keil, IAR,
STM32CubeIDE) based on STEVAL-STWINKT1B

• ST BLE Sensor Application for Android/iOS to download from Google Store / iTunes

• ST DSH-PREDMNT Account

FP-AI-PREDMNT2 is not the default firmware on STEVAL-STWINKT1B.

To update the firmware, please follow the instructions available in slide 11

10
 Setup & Demo Examples
Firmware update

FP-AI-PREDMNT2 is not the default firmware on the STEVAL-

STWINKT1B, so it needs to be downloaded on the board by the user.
The easiest way is to use the pre-compiled binary provided in the
package in the folder Projects\STM32L4R9ZI-
STWIN\Applications\PREDMNT2\Binary.
To update the firmware:
• Connect the STWIN core system board to the STLINK-V3MINI
programmer.
• Connect both the boards to a PC using micro USB cables.
• Open STM32CubeProgrammer, select the proper binary file and
download the firmware.

For further details, see UM2937

STLink-V3MINI
11
 FP-AI-PREDMNT2
Setup Overview

1 www.st.com/stm32ode-fp FP-AI-PREDMNT2 package structure

Docs
2 BSP, HAL,
3 drivers
BlueNRG,
Download & unpack FreeRTOS,
AI-Library, …
Select Function Pack: FP-AI-PREDMNT2 Applications
4
Configure the board through ST BLESensor
app and monitor the application through the
ST DSH-PREDMNT dashboard
6 Projects\STM32L4R9ZI-STWIN\Applications\PREDMNT2\

12
 FP-AI-PREDMNT2
Setup Overview
STWIN Power ON/OFF

• Battery only (no USB cable): PWR

• Power ON
• Long-press the PWR button until the red led turns off (~1 sec)
• Power OFF
• Press the PWR button

• Plugged mode (USB cable)

• Power ON
• When USB is plugged-in, the STWIN is always on. It
doesn’t matter if the battery is present or not
• Power OFF
• Unplug the cable and, if the battery is connected, press the
PWR button.

13
2.1- Setup wi-fi and security certificates
through ST BLESensor app
 Wi-fi configuration and certificate request
• Once the firmware has been loaded, press the [Reset] black button on the STWIN core system. Then, open
STBLESensor app and connect to PM2_100 device.
• Open the menu, select [Board configuration] and open [Certificate Request].
• Select [Predictive Maintenance] and sign in with your MyST credentials.
 Wi-fi configuration and certificate request
• Ask for a new certificate by tapping the + symbol placed at the bottom of the screen or retrieve the ones
already associated to your myST account. The certificates valid for your STWIN are highlighted in blue
• Tap on the key to download the security certificates into the STWIN and tap on the Wi-Fi symbol to update Wi-
Fi credentials.

Once the procedure ends correctly, the orange LED switches on.
16
Now the STWIN is connected to the AWS PREDMNT dashboard and is ready to stream data.
2.2- Setup and use DSH-PREDMNT
dashboard
How to use DSH-PREDMNT dashboard
DSH-PREDMNT features

18
 How to use DSH-PREDMNT dashboard
Configure periodic update time

• Set streaming times to reduce bandwidth according to the system dynamics

1 2

Drag balloons

Configure transmission
periods

19
 How to use DSH-PREDMNT dashboard
Configure measures and condition monitoring thresholds

• Configure the measures to take under control and set thresholds for warning and alerts
1 2 3
Threshold
configuration

Configure alarms

20

Measures domains
 How to use DSH-PREDMNT dashboard
Add a device to live monitoring

• Once devices are selected in

the live dashbaord, a new
cardbox will be showed and,
if your physical device is
transmitting, you will see live
data monitoring refreshing.
• Data showed are customized
according to the FW release
of the device, they are:
• Environmental Analysis
• Vibration Analysis in Time
Domain
• Vibration Analysis in Frequency
Domain
Select your device
• Sound Emission Analysis for live monitoring

21
 How to use DSH-PREDMNT dashboard
Hand Shaking Device – Flexible visualization

• The Dashboard recognizes the device, update the device shadow (that will be used for other
purposes also) and adapt the visualization

Vibration (TDM)
Environmental domain
domain

Vibration (FFT) domain

Acoustic domain

22
 FP-AI-PREDMNT2
DHS-PREDMNT dashboard to visualize data and setup STWIN

23
2.3- Setup NanoEdge™ AI library by
Cartesiam

24
 Setup a NanoEdge™ AI library into FP-AI-PREDMNT2
Build library using Embed into
Generating contextual data using Use with
NanoEdge™ AI FP-AI-
FP-SNS-DATALOG1 DSH-PREDMNT
Studio PREDMNT2

Build Embed Process

Capture data Label data
NanoEdge Lib NanoEdge Lib new data

NanoEdge
Normal Modeling
integration Real-time
Accelerometer
NanoEdge learning
Anomaly Results
implementation

1 2 3 4 5
25
 Capture data and create a labelled dataset
Generating contextual data using FP-SNS-DATALOG1

• Flash STWIN board with

HSDatalog firmware
• Acquire training data onto SDCard
• Extract and convert logs using
hsdatalog_to_cartesiam.py Python
1 Generate data set 2 Label Data
script
Data log
Regular signals file.csv

Abnormal signals file.csv

SDCard

normalDataFull.csv

abnormalDataFull.csv

Serial
 Use NanoEdge AI Studio tool
Build library using NanoEdge™ AI Studio

• Generate a Machine Learning

library with NanoEdge™ AI Studio
based on the provided data 3 Building and Export Library
• Download the library from
the NanoEdge™ AI Studio
file

1. Libneai.a
2. NanoEdgeAI.h
27
 Integration in the FP
Embed into FP-AI-PREDMNT2

• Replace the provided stub in the

function pack by the newly
generated library
• Rebuild the firmware
4 Integrate Lib files into the
• Program the STWIN with the new
firmware project

NanoEdgeAI.h
Libneai.a

5 Re-compile the FP
 Running and Learning
Use FP-AI-PREDMNT2 with DSH-PREDMNT

• Learn the normal modes on the

edge
6 NanoEdge 7 Run the demo: NanoEdgeAI
Learn execute

NanoEdgeAI_learn() NanoEdgeAI_execute()
29
 Running and Re-training
Use FP-AI-PREDMNT2 with DSH-PREDMNT

• Monitor anomalies in detection

• Go back to learn mode to improve
the learning or to add new normal
7 Analyze and decide if new 8 Run the demo
modes
training should be launched again

30
 Send a command to NanoEdge™ AI library
A
• DSH-PREDMNT allow you to configure
and setup your custom NanoEdge™ AI
library
A. Click on the Configuration tab
B. Select the symbol of the desired
STWIN board
C. Click on Anomaly detection
B

C
 Send a command to NanoEdge™ AI library

D E

• The settings tab allow to setup the sensitivity and the threshold parameters
• The commands tab allow to chose either train or detect command. The result of the
train command is the training of the neai library for the time interval set in the “phase
duration” form. The detect command will trigger the anomaly detection phase in the
application in according with the generated library
 Detection phase of NanoEdge™ AI library

2
1 Orange led is blinking

Start detection

Alarm sent

• If the detect command is set, in according with the custom NanoEdge™ AI library, an
anomaly event is sent to the dashboard and an event will be shown in the EVENTS
panel of the dashboard.
• For further details on how NanoEdge™ AI libraries work, read the
detailed documentation of NanoEdge™ AI Studio.
3- Documents & Related Resources
 Documents & Related Resources
All documents are available in the DESIGN tab of the related products webpage

FP-AI-PREDMNT2:
• DB4566: STM32Cube function pack for STEVAL-STWINKT1B evaluation kit for predictive maintenance application based on
artificial intelligence (AI) – databrief
• UM2937: Getting started with the STM32Cube function pack for STEVAL-STWINKT1B evaluation kit for predictive
maintenance application based on artificial intelligence (AI) – user manual
• Software setup file
STEVAL-STWINKT1B:
• Gerber files, BOM, Schematic
• DB4345: STWIN SensorTile Wireless Industrial Node development kit and reference design for industrial IoT applications –
databrief
• UM2777: How to use the STEVAL-STWINKT1B SensorTile Wireless Industrial Node for condition monitoring and predictive
maintenance applications – user manual
STEVAL-STWINWFV1:
• Gerber files, BOM, Schematic
• DB3971: Wi-Fi expansion for the SensorTile Wireless Industrial Node (STWIN) kit – databrief

Consult www.st.com for the complete list 35

 Documents & Related Resources
All documents are available in the DESIGN tab of the related products webpage

DSH-PREDMNT:
• Cloud based web application for condition monitoring and predictive maintenance

To build up your custom NanoEdge AI library:

NanoEdge AI Studio:
• Automated Machine Learning (ML) tool for STM32 developers
FP-SNS-DATALOG1:
• STM32Cube High Speed Datalog function pack for STWIN evaluation kits

Consult www.st.com for the complete list 36

4- STM32 Open Development
Environment: Overview
 STM32 ODE Ecosystem
FAST, AFFORDABLE PROTOTYPING AND
DEVELOPMENT
The STM32 Open Development Environment (ODE) is an open, flexible, easy and
affordable way to develop innovative devices and applications based on the STM32 32-bit
microcontroller family combined with other state-of-the-art ST components connected via
expansion boards. It enables fast prototyping with leading-edge components that can quickly
be transformed into final designs.
The STM32 ODE includes the following five elements:
o STM32 Nucleo development boards. A comprehensive range of affordable development
boards for all STM32 microcontroller series, with unlimited unified expansion capability, and
with integrated debugger/programmer
o STM32 Nucleo expansion boards. Boards with additional functionality to add sensing,
control, connectivity, power, audio or other functions as needed. The expansion boards are
plugged on top of the STM32 Nucleo development boards. More complex functionalities STM32 Nucleo STM32Cube
can be achieved by stacking additional expansion boards development boards development boards

o STM32Cube software. A set of free-of-charge tools and embedded software bricks to

enable fast and easy development on the STM32, including a Hardware Abstraction Layer, STM32 Nucleo STM32Cube
middleware and the STM32CubeMX PC-based configurator and code generator expansion boards (X-NUCLEO) expansion software (X-CUBE)

o STM32Cube expansion software. Expansion software provided free of charge for use with
STM32 Nucleo expansion boards, and compatible with the STM32Cube software
framework Function Packs

o STM32Cube Function Packs. Set of function examples for some of the most common
application cases built by leveraging the modularity and interoperability of STM32 Nucleo
development boards and expansions, with STM32Cube software and expansions.
The STM32 Open Development Environment is compatible with a number of IDEs including
IAR EWARM, Keil MDK, mbed and GCC-based environments.
 STM32 Open Development Environment:
all that you need
The combination of a broad range of expandable boards based The building blocks Your need Our answer
on leading-edge commercial products and modular software,
from driver to application level, enables fast prototyping of ideas Accelerometer, gyroscope
Inertial modules, magnetometer COLLECT
that can be smoothly transformed into final designs. Pressure, temperature, humidity
Sense
Proximity, microphone

To start your design: Bluetooth LE

TRANSMIT
Sub-GHz radio
Connect
•
NFC, GNSS
Choose the appropriate STM32 Nucleo development board (MCU) IO-Link

and expansion (X-NUCLEO) boards (sensors, connectivity, audio,

motor control etc.) for the functionality you need Touch controller
HMI
ACCESS
LED drivers

• Select your development environment (IAR EWARM, Keil MDK, and

GCC-based IDEs) and use the free STM32Cube tools and software. Stepper motor driver ACTUATE
DC & BLDC motor driver Move
• Download all the necessary software to run the functionality on the
selected STM32 Nucleo expansion boards.
Industrial input / output POWER
• Compile your design and upload it to the STM32 Nucleo Energy management & battery
USB
Power
development board.
• Then start developing and testing your application. General-purpose microcontrollers
Process PROCESS
Secure microcontrollers

Software developed on the STM32 Open Development

Environment prototyping hardware can be directly used in an Software
advanced prototyping board or in and end product design
using the same commercial ST components, or components
from the same family as those found on the STM32 Nucleo Power supply
through USB
or external
STM32
microcontroller

boards. source

Integrated Complete product range

debugging and
programming from ultra-low power to
ST-LINK probe high-performance
ST morpho extension header

Arduino™ UNO R3 extension

headers
39
 Thank you

© STMicroelectronics - All rights reserved.

The STMicroelectronics corporate logo is a registered trademark of the STMicroelectronics
group of companies. All other names are the property of their respective owners.