 Objects are generating a ton of data and processing them allows our customers to add iron feature to their products. With the growth of the IoT, our customers are always looking more effective at communication protocol and that's true with the STM42W series, our MCRs are bringing efficient wireless connectivity. However, we are convinced that sending more data faster is not the only way to tackle this problem. Today, I will show you how making your product smarter will allow you to save power consumption, to increase security and privacy and to reduce latency, to provide better user experience. Adding attention to your product will allow you to optimize the way you are managing your bandwidth. For example, adding artificial intelligence at the edge on your STM42, such as computer vision will allow you to process locally good data and then send only relevant values instead of big images. The demo behind me is illustrating perfectly the example I previously mentioned. For example, the data is read thanks to a low-cost camera mounted on the water meter and the first AI model is detecting the region of interest. So where are the digits on the images and the second neural network is recognizing the character in order to send only the numerical value to the dashboard through CIRA connectivity? Here on this demo, we are sending only 5 bytes for the value instead of almost 50 kilobytes to send the raw images through connectivity. Both models are running locally on the ultra low-power STM42. Using edge processing can be applied to a wide range of areas. For example, smart factory with predictive maintenance, smart automation for asset tracking with activity and reconnection, or smart building, smart home with people recognition, people tracking, and phase 3 code and way more. Hasumain or machine learning can take a wide range of pain points and you will be able to find more information on our website, st.com slash STM42 AI.