Schneider Electric and Swiss electronics company ST Microelectronics are demonstrating a prototype IoT sensor that enables building-management services and efficiency gains by understanding building-occupancy levels and usage.

The two companies have collaborated to integrate artificial intelligence (AI) into a people-counting sensor, which overcomes the problem of monitoring attendance in large spaces with multiple entrance points.  

With the digitisation of building occupancy, French firm Schneider is following its mission to be a digital partner for sustainability and efficiency by delivering insights such as queue monitoring to assist smart building management while respecting individuals’ privacy by design. The IoT sensor has been developed by combining the expertise of STM’s AI group and the sensor-application expertise of Schneider Electric to identify and embed an object-detection neural network in a small microcontroller (MCU).

“This promising technology opens new attendance monitoring and people counting in numerous applications such as monitoring queues, building usage and social distancing,” said Maxime Loidreau, IoT sensors programme manager at Schneider Electric. “Our innovative demonstration, created with ST Microelectronics, finds applications in various segments, from hotels to offices and retail, and more generally any building where knowing attendance and space occupation has a value. This will redefine the building of the future.”

The increase in design productivity comes from its use of the STM32Cube AI toolchain, which has capabilities for developing AI applications for the STM32 MCUs. This allowed Schneider to gain flexibility and efficiency in hardware design from the engineering resources, sophistication and ease of use provided by the STM32Cube software-development ecosystem.

The prototype people-counting sensor combines a Lynred ThermEye thermal imager integrated in a low-power design created by Schneider, with a Yolo-based neural network model running on the recently introduced STM32H723 MCU.

“This project demonstrates the power of deep learning to enhance embedded data-processing performance, showing how high-value applications can be hosted on a cost-effective microcontroller-based platform,” said Miguel Castro, AI business line manager at ST Microelectronics. “Our STM32Cube AI ecosystem empowers users to create flexibility within a fast time-to-market window. Customers can enjoy even greater productivity leveraging the support of our technical team to overcome engineering challenges.”

The STM32 AI ecosystem provides building blocks for neural networks to run on STM32 MCUs. Various deep-learning frameworks such as Keras, TensorFlow Lite and Onnx exchange format are supported natively.

Included in the ecosystem is the X-Cube-AI software expansion package, which extends the capabilities of the STM32Cube MX initialisation tool to convert automatically pre-trained neural networks, generate optimised libraries for the target MCU, and integrate these into the user’s project. Additional support to automate laborious development tasks includes several ways of validating neural network models and measuring performance on STM32 MCUs without creating the necessary C code by hand.

The general DNN approach supported by STM’s software-development ecosystem, mapped onto the STM32 portfolio, lets users efficiently replicate development effort to create products for multiple markets. The STM32H723 MCU powering the demonstration has credentials for hosting AI applications, including high core performance, up to 1Mbyte flash, high-speed off-chip memory interfaces, and integrated features for connecting various sensor types.