The Smart Connected Sensors platform from Bosch Sensortec enables high-performance full-body movement tracking via a wearable reference design consisting of 8x sensor nodes, with an accompanying SDK and AI-capable software toolkit. It supports fast development and time to market for applications that include fitness, activity tracking, and physical rehabilitation, but also wearables and hearables for Augmented Reality (AR), Virtual Reality (VR), and gaming.
Each wearable sensor node has a 36 x 22.2 x 10-mm wristband-type form factor (strap not included) and is connected via Bluetooth 5.3 LE. At 25 Hz data rate (max. 50 Hz), this enables a battery lifetime of up to 40 hours, thanks to smart power management within the device. The following sensors are used to track the motion of the user:
- BHI380 6-axis smart IMU with AI functionality for movement tracking
- BMP581 barometric pressure sensor for accurate relative vertical distance measurement
- BMM350 magnetometer for additional sensing degrees of freedom
This combination of sensors in conjunction with the software provides accurate full-body motion mapping onto virtual avatars without the requirement for cables, bodysuits, or specialist cameras. Synchronisation is down to 1 millisecond over 24 hours for extremely precise tracking. This design is certified for use in all regions, and Bosch supplies all hardware design files for reference during development.
The key sensing component within this platform is Bosch’s BHI380 smart IMU, which integrates a low-power 32-bit MCU with an AI-capable ARC EM4 CPU running up to 3.6 CoreMark/MHz. Suitable for the most compact designs, the BHI380 is available in a package of 2.5 x 3 x 0.95 mm. The standby current is 8 µA, and this increases to <600 µA in run mode, where activity recognition with feedback is enabled.
Preinstalled sensor fusion software and AI algorithms enable improved gesture recognition across many applications. In physical rehabilitation, for example, feedback can be fed to the Smart Connected Sensors software to inform patients whether their movements match what is required for treatment. Multi-device-gesture recognition is possible with minimal code, with a base code of 39 kB and an average complex pattern size of 2 kB.
Learn more about Bosch’s Smart Connected Sensors platform in this official product flyer, and if you have a commercial project and want to put this technology or the BHI380 to the test, fill out the form below, and ipXchange will get you connected.