Nowadays there is a growing need for medical devices to provide more precise and accurate monitoring of vital signs while also extending their battery life to enhance their overall efficiency. Achieving these objectives will require the integration of advanced technologies, such as machine learning, artificial intelligence, and sensor technology, to improve the accuracy and reliability of medical devices. By optimizing the performance of medical devices, healthcare professionals can provide more effective care to patients, which can ultimately lead to better health outcomes. Therefore, it is imperative that medical device manufacturers continue to innovate and develop devices that incorporate the latest technologies and meet the evolving needs of the healthcare sector.
In-Sensor AI tech revolutionizes edge and always-on devices, conserving battery by processing data directly on the sensor and transmitting only relevant info to the main MCU.
We will present a practical joint project with STMicroelectronics, featuring ultra-compact Neuton Neural Networks integrated into next-gen resource-limited sensors, including the STMicro ultra-low-power 6-axis ISM330IS ISPU sensor. This sensor has 8KB of Data RAM and 32KB of Program RAM.
Advantages of In-Sensor AI Architecture:
This technology optimizes the energy consumption of the entire solution. Instead of sending raw data from the sensor to the MCU, processed data is sent, allowing the MCU to be in sleep mode. In-Sensor AI architecture reduces battery consumption since there is no need to wake the host MCU to read and process sensor data. All machine learning analysis is performed inside the low-power sensor. It unloads the MCU bandwidth by freeing up the CPU from algorithm processing and thread switching, critical for hard real-time systems. Additionally, it simplifies software development since the introduction of distributed systems streamlines the software design process, requiring significantly less time for multithreading and managing shared MCU resources.
During the presentation, we will provide an overview of applications and an example of In-Sensor AI implementation to demonstrate a tracking solution for toothbrushes. This solution identifies 15 zones of the oral cavity in real-time directly on the device: • top left front • top left back • top left biting area • bottom left front • bottom left back • bottom left biting area • top right front • top right back • top right biting area • bottom right front • bottom right back • bottom right biting area • front teeth front side • front teeth top back side • front teeth bottom back side with a total footprint of around 13kB.
Conclusion:
In-Sensor AI is an emerging technology that is gaining traction and being successfully adopted in various industries, especially in MedTech. Edge devices equipped with programmable sensors are becoming increasingly popular among consumers due to their advanced features. The potential applications of this technology are vast and diverse, and their energy efficiency is a significant advantage, making them a perfect fit in scenarios where battery life is critical.
