
Bosch BHI260AP Self-Learning AI Smart Sensor
Bosch BHI260AP Self-Learning AI Smart Sensor with Integrated Inertial Measurement Unit (IMU) is an all-in-one solution for always-on sensor applications, including fitness tracking, navigation, machine learning analytics, and orientation estimation. The sensor features a variety of software functionalities, 32-bit customer-programmable microcontroller, and a six-axis IMU. The BHI260AP supports self-learning AI software for fitness tracking, swim analytics, pedestrian dead reckoning, and relative and absolute orientation. Bosch BHI260AP Self-Learning AI Smart Sensor with Integrated IMU is ideal for use in wrist wearables, head-mounted devices, smartphones, and AR / VR / MR headsets and controller devices.The Bosch BHI260AP Self-Learning AI Smart Sensor is housed in a compact 3.6mm x 4.1mm x 0.83 LGA-44 package, ideal for space-constrained applications.
Features
- CPU core
- ARC EM4 CPU with ARCv2 16-/32-bit instruction set (up to 3.6 CoreMark/Mhz)
- Floating point unit (FPU) / memory protection unit (MPU)
- 4-channel micro DMA controller / 2-way associative cache controller
- Integrated sensor (6-degrees-of-freedom [DoF] IMU)
- 16-bit 3-axis accelerometer
- 16-bit 3-axis gyroscope
- Self-learning AI software for fitness tracking, swim analytics, pedestrian dead reckoning, relative and absolute orientation
Applications
- Wrist wearables
- Smartwatches
- Fitness bands
- Smart hybrid watches
- Smartphones and other mobile communication devices
- Head-mounted devices
- Headsets
- Truly wireless in-ear devices
- Smart sunglasses
- AR / VR / MR headset and controller devices
Videos
Block Diagram

Connection Diagram

Notes:
- Recommendation: Connect to VDDIO if not needed
- Leave open if not needed
- When using host interface in I2C mode: connect R1...R3 as indicated; Do not apply host connection to HCSB and HSDO; HSDO can be used for device address selection if needed
- Host signal name depends on SPI/I2C configuration
- Example configuration shows one external I2C bus on M3. Different configurations, e.g. using Ml and/or M2, are also possible
Published: 2020-11-04
| Updated: 2023-05-08