Arrow Electronics today unveiled a new AI thermal sensing solution designed to help engineers and product designers accelerate the development of smart, reliable, and affordable health-monitoring devices.
There is demand for contactless, robust and privacy-preserving devices for detecting temperature with accuracy and wide coverage to safeguard against public health emergencies. According to a recent research report, the global thermal imaging market is expected to reach USD 4.6 billion by 2025, growing at a CAGR of 6.2%1.
The AI thermal sensing solution powered by STMicroelectronics’ X-CUBE AI optimises bill-of-materials and simplifies hardware and software integration, making it a good choice to deploy for temperature-screening devices and a wide variety of other consumer-grade and healthcare applications.
"Many product engineers find it challenging to deploy AI models in a cost-efficient and robust way to a wide range of consumer-grade, healthcare, and IoT edge devices. This AI-enhanced solution with STM32 Arm Cortex-M-based microcontrollers offers a pricing advantage and delivers better energy efficiency, making it easier to bring to market the kind of intelligent devices that help individuals and communities stay safe," said Jacky Wan, Arrow's vice president of engineering for APAC.
Arrow's AI-powered thermal sensing solution can achieve a quick and accurate temperature screening, with multiple individuals screened simultaneously. The integrated solution is comprised of four blocks:
- A STMicroelectronics Time-of-Flight (ToF) sensor which allows absolute distance measurement whatever the target colour and reflectance, providing accurate ranging up to 400 cm and speed up to 60 Hz
- A STMicroelectronics digital ambient thermal sensor which detects ambient temperature and dynamically compensates for differences to allow complex measurements at high speed
- A longwave infrared thermal image sensor developed and fabricated by Meridian Innovation, a Hong Kong-based pioneering developer of advanced CMOS technology
- STMicroelectronics' dual-core M7 and M4 MCU running at 480Mhz which acts as the main processor of the system
A thermal image noise reduction process making use of AI deep learning is deployed in the system. This special denoise process runs on a convolutional neural network (CNN). With the STMicroelectronics' AI expansion pack - STM32Cube.AI, the deep-learning algorithm can be exported and executed on STM32 Arm Cortex-M-based microcontrollers. The solution can detect target object distance and human presence. Images can be displayed in heat map or RGB format.
To provide the best performance, an ultra-low noise LDO with a noise voltage of only 6 micro volt RMS is used for the power supplier of all the sensors. All three sensor blocks are connected to the main processor using an I2C and SPI bus, which will handle processing and calculation of the environmental data collected by the sensors to increase system accuracy.
Adding the wireless connectivity feature for transmitting data to the gateway/cloud, the integrated solution can incorporate Murata type 1YN, a small and high-performance module with Single-Band 2.4GHz Wi-Fi 802.11n and Bluetooth 5.2.
Matteo Maravita, head of APAC AI Competence Centre of STMicroelectronics, said the Arrow team has done a great job in developing this integrated solution.
“This solution takes full advantage of AI techniques on STMicroelectronics microcontrollers and sensors thanks to STM32Cube.AI, a complete development ecosystem from STMicroelectronics that allows easy implementation of complex AI models onto STMicroelectronics products,” Maravita said.
“The integrated solution allows product engineers to rapidly integrate AI and thermal-sensing technology and accelerate the product development cycle of smart, reliable and low-cost temperature-sensing devices, addressing the consumer-grade and healthcare market needs," he added.