Slide 1: Enhance your products using edge AI solutions on STM32
Enhance your products using edge AI solutions on STM32
Slide 2: A short introduction to AI concepts
A short introduction to AI concepts
Slide 3: What is artificial intelligence
What is artificial intelligence Embedded artificial intelligence The ability of machines to perform tasks From a technical perspective, AI is that typically require human intelligence another way to program a machine. Perception Reasoning Learning Decision-making
Slide 4: Using AI algorithms implies two phases
Using AI algorithms implies two phases Training Inference AI models are produced using datasets and a Then during the second phase, we can use this model. training engine/framework. This is the inference phase. ST Sensor
Slide 5: Embedding intelligence at the edge
Embedding intelligence at the edge
Slide 6: Inference shift from the cloud to the tiny edge
The explosion of AI-enabled devices is accelerating the inference shift from the cloud to the tiny edge Inference Cloud-centric On device-centric Tiny edge-centric Enabled by a different class of hardware and software
Slide 7: Inferring at the edge brings substantial benefits
Inferring at the edge brings substantial benefits This enables AI Ultra-low latency Real-time applications Tiny edge-centric 01 Reduced data transmission 10 Generate meaningful information Enhanced privacy and security No data data sharing sharingininthe thecloud cloud Sustainable on energy Low data
Slide 8: Opening a new range of embedded AI applications
Opening a new range of embedded AI applications Upgrade existing devices Reduce BoM by shifting from Offer advanced services with AI-based services MPUs to MCUs without cloud costs Arc Fault People Object Detection Detection Classification Predictive Sound Pose Maintenance Analysis Estimation Batter
Slide 9: Edge AI application examples
Edge AI application examples
Slide 10: AI addresses a wide variety of projects
AI addresses a wide variety of projects Smart Smart city Smart home Energy Industry 4.0 Healthcare buildings Energy optimization Energy optimization Outdoor cameras Solar panels Predictive maintenance Wearables Traffic management Access control Babycams Breaker's control Tools safety Patient monitor
Slide 11: Washing machine using advanced motor control algorithms
A washing machine uses advanced motor control algorithms to weigh clothes and optimize water, detergent, and energy used ~15-40% Leader in white-goods Energy saving per Production starting in 2024 washing cycle for millions units
Slide 12: Virtual sensor: tire pressure detection on e-bikes
Adding tire pressure detection capability to e-bikes without adding any new hardware Virtual sensor Tire pressure measured Panasonic e-bike through the e-motor Hundreds of thousands of current consumption units annually, starting in 2024
Slide 13: Predictive maintenance
After-market predictive maintenance intelligent sensor with wireless connectivity Predictive maintenance EMEA company Multisensors and Deployed at Volvo Trucks learning on device manufacturing plant
Slide 14: Explore inspiring case studies
Explore inspiring case studies Gain valuable insights from real-world case studies on how to improve your embedded solutions using edge AI. Edge AI case studies Edge AI case studies
Slide 15: ST's investment in edge AI enabling broad adoption
ST’s investment in edge AI enabling broad adoption
Slide 16: More than a decade of research, development, and deployment
More than a decade of research, development, and deployment ST Neural-ART accelerator in STM32N6 Research using Edge AI on Edge AI ML techniques for Automotive MCUs Developer Cloud visual search Edge AI for sensors (MLC) Introduction of AutoML Edge AI devices 2012 2015 2018 2019 2020 2021 2022 2023
Slide 17: The challenge of deploying embedded AI
The challenge of deploying embedded AI Memory footprint AI expertise Inference time eSW development Data Power consumption Deploying embedding AI on microcontrollers presents significant challenges
Slide 18: Taking up the edge AI developer's challenge
Taking up the edge AI developer’s challenge Performance Create, deploy Security & maintain ML models Power efficiency
Slide 19: A comprehensive approach to accelerate product transformation
A comprehensive approach to help developers accelerate their product transformation NEW ST Edge AI Suite STM32 MCUs STM32N6 MCU NEW Hardware Software platforms tools STM32 MPUs STM32MP2 MPU Stellar MCUs MEMS MLC/ISPU Click on the products to find out more
Slide 20: STM32 product offering
STM32 product offering SENSING (TIME SERIES) AUDIO VISION Enabling major edge AI technologies HORSE + DOG + CAT ANOMALY DETECTION SENSING SENSING SENSING Software AD AUDIO AUDIO tools for any VISION VISION X-LINUX-AI user profile User-friendly AutoML tool AI model optimizer and code A complete AI fr
Slide 21: ST Neural-ART Accelerator in STM32N6
ST Neural-ART Accelerator in STM32N6 ST proprietary ML performance uplift* neural processing unit (NPU) Specialized hardware accelerator performs advanced AI inference Achieving edge AI performance within MCU power budget Read more in the blog Read our whitepaper 21 * 600 GOPS NPU vs 1 GOPS NN peak
Slide 22: How the STM32N6 changes the game
How the STM32N6 changes the game An MPU-like end-user experience. Available on an MCU. People detection Pose estimation Hand landmark A blue play button with a black background Description automatically generated Watch all videos on YouTube Watch all videos on YouTube 22
Slide 23: Software ecosystem
Jumpstart your project with an STM32 MCU 7 _ Follow these few simple steps to get started. Microcontrollers, boards and hardware tools, software tools and embedded software packages, and so much ed 3,300+ 500+ 135,000+ pertnumbers. software & boards, members
Slide 24: A broad collection of free tools
A broad collection of free tools DATA Bring your own data (BYOD) MODEL MODEL ZOO ML Bring your own model (BYOM) ML ML ML ML AI for OpenSTLinux Neural network model optimizer Neural network model optimizer & app examples Neural network model optimizer & benchmark Automated machine learning STM32Cube.
Slide 25: NanoEdge AI Studio, the AutoML tool
NanoEdge AI Studio, the AutoML tool
Slide 26: Software tool: NanoEdge AI Studio
Software tool: NanoEdge AI Studio DATA Bring your own data (BYOD) MODEL MODEL ZOO ML Bring your own model (BYOM) ML ML ML ML AI for OpenSTLinux Neural network model optimizer Neural network model optimizer & app examples Neural network model optimizer & benchmark Automated machine learning STM32Cube
Slide 27: Simplified edge AI development workflow
Simplified edge AI development workflow Deployment of NanoEdge AI Studio libraries, the market reference AutoML tool, is completely free for unlimited quantities on any STM32 … and available on any Arm® Cortex® -M MCU* The best combination for given data: ML model, hyperparameters, and preprocessing
Slide 28: State-of-the-art machine learning for smarter products
State-of-the-art machine learning for smarter products I want to anticipate product failures I need to detect any outliers I want to identify the activity, the environment, the usage I need to predict future states 28
Slide 29: NanoEdge AI Studio workflow
NanoEdge AI Studio workflow Create your dataset Analyze, choose, and validate Generate an Train your model And set up your project the best ML processing optimized library directly on the device Project parameters 011001 max ram, max flash, 100111 microcontroller ref. 110100 .a Library 111000 011001
Slide 30: All the tools needed for data preparation
All the tools needed for data preparation I want to clean my data quickly and efficiently” 30
Slide 31: Get the best ML algorithm for your application
Get the best ML algorithm for your application Internal benchmarking tool picks the best algorithm for your data NanoEdge AI Studio And optimizes it to reduce improves the performance footprint and latency of the model. 31
Slide 32: STM32Cube.AI & ST Edge AI Developer Cloud
STM32Cube.AI & ST Edge AI Developer Cloud
Slide 33: Software tool: STM32Cube.AI
Software tool: STM32Cube.AI DATA Bring your own data (BYOD) MODEL MODEL ZOO ML Bring your own model (BYOM) ML ML ML ML AI for OpenSTLinux Neural network model optimizer Neural network model optimizer & app examples Neural network model optimizer & benchmark Automated machine learning STM32Cube.AI fo
Slide 34: One tool – two versions to deploy AI on STM32
One tool – two versions to deploy AI on STM32 Generate ST Edge AI model zoo Optimize your NN model optimized code OR Bring your own model STM32Cube.AI for desktop STM32Cube Command-line ecosystem interface ST Edge AI Developer Cloud via via Optimized model Online REST Benchmarking tool platform API
Slide 35: The pillars of STM32Cube.AI
The 3 pillars of STM32Cube.AI Graph optimizer Quantized model support Memory optimizer Automatically improve performance through Import your quantized ANN to be compatible Optimize memory allocation to get the best graph simplifications & optimizations that with STM32 embedded architectures while pe
Slide 36: Start with edge AI optimized models: STM32 model zoo
Start with edge AI optimized models STM32 model zoo A collection of application-oriented models optimized for STM32 Human activity Image classification recognition Hosted on GitHub Motion Sensing Computer vision <I> Model training scripts • Scripts to generate and validate Audio event detection Obje
Slide 37: Discover the ST Edge AI Developer Cloud
Discover the ST Edge AI Developer Cloud 37
Slide 38: ST Edge AI Suite
ST Edge AI Suite
Slide 39: 50+ case studies
By bringing solutions to engineers and data scientists at every stage of their development, the ST Edge AI Suite accelerates edge AI adoption. 50+ case studies 10+ free software tools Unified AI core technology
Slide 40: Access to enabling hardware and software for edge AI
Access to the enabling hardware and software for running edge AI on MCUs, MPUs, and smart sensors. Empowering developers, data scientists, and product designers to harness the power of edge AI. A digital experience available in English, Chinese, and Japanese.
Slide 41: Explore inspiring case studies
Explore inspiring case studies Gain valuable insights from real-world case studies on how to improve your embedded solutions using edge AI. Edge AI case studies Edge AI case studies
Slide 42: Free tools to run edge AI on MCUs, MPUs, and smart sensors
Free tools to run edge AI on MCUs, MPUs, and smart sensors Find the tools you need to optimize and deploy machine learning algorithms, from data collection to final deployment on hardware. Bring your Bring your OR own data own model AutoML Model Ideation Data Benchmark Deploy End-to-end flow optimiz
Slide 43: Find out more at www.st.com/stm32ai
Find out more at www.st.com/stm32ai © STMicroelectronics - All rights reserved. ST logo is a trademark or a registered trademark of STMicroelectronics International NV or its affiliates in the EU and/or other countries. www.st.com/trademarks For additional information about ST trademarks, please ref