0
Robotics and extended reality share a common demand: the motion sensor must perform flawlessly when the platform is moving fast, accelerating hard, and vibrating under load. A slow, gentle gesture is easy to measure; a robot arm executing a rapid trajectory or an XR controller swung in a combat game is a different problem entirely. The BMI563 is Bosch Sensortec’s response to this class of high-dynamic applications. As the extended-range variant of the BMI5 platform, it combines a ±4,000 dps gyroscope and ±32 g accelerometer with the platform’s vibration robustness and Edge-AI capabilities. This article examines the BMI563 as an independent design choice for robotic SLAM, XR controller tracking, and action camera stabilization.
The most visible differentiator of the BMI563 is its full-scale range. The ±4,000 dps gyroscope and ±32 g accelerometer represent a doubling of the previous generation’s range, ensuring saturation-free data acquisition even in the upper measurement range. For an XR controller, this means that a fast swing or a sudden impact does not clip the gyroscope output, preserving the continuity of orientation tracking that is essential for a stable virtual experience. For a robot operating near its dynamic limits, the extended range captures the full motion profile without the data loss that would occur with a narrower-range sensor.
The BMI563 inherits the BMI5 platform’s vibration robustness, maintaining accurate measurement across a frequency range from 0 to 40 kHz and into the MHz range. In robotics, vibration originates from motors, gearboxes, and ground contact; in XR controllers, it comes from haptic actuators and the user’s grip. The sensor’s ability to reject these disturbances while preserving the signal of interest is what enables reliable SLAM and motion tracking. The noise density—below 50 µg/√Hz for the accelerometer and below 3 mdps/√Hz for the gyroscope—provides the low noise floor necessary to resolve the subtle movements that accompany the large ones, such as the fine stabilization corrections in a robot’s end effector or the gentle aim adjustments in an XR application.
|
Attribute |
BMI563 Specification |
|
Package Dimensions |
2.5 × 3.0 × 0.75 mm³ LGA |
|
Accelerometer Ranges |
±2, ±4, ±8, ±16, ±32 g |
|
Gyroscope Ranges |
±125, ±250, ±500, ±1000, ±2000, ±4000 °/s |
|
Accelerometer Resolution |
16-bit & 24-bit |
|
Accelerometer Noise Density |
< 50 µg/√Hz (≤8 g); < 60 µg/√Hz (16 g) |
|
Gyroscope Noise Density |
< 0.003 °/s/√Hz |
|
Accelerometer ODR |
1.56 Hz to 6.4 kHz |
|
Gyroscope ODR |
12.5 Hz to 12.8 kHz |
|
Current (Max ODR) |
650 µA |
|
Current (100 Hz ODR) |
250 µA |
|
Suspend Current |
3 µA |
|
Primary Interface |
MIPI I3C®, I²C, SPI |
|
FIFO |
8 KB |
|
Supply (VDD / VDDIO) |
1.71–3.6 V / 1.08–3.6 V |
|
Temperature Range |
–40°C to +85°C |
The BMI563 includes the same programmable Edge-AI classification engine found across the BMI5 platform. In a robotics context, this engine can be configured to recognize specific motion signatures—a collision event, a stall condition, or a particular trajectory segment—and signal the host only when intervention is required. In an XR controller, it can classify gestures or detect when the controller is at rest, allowing the host to enter a low-power state without sacrificing responsiveness. By analyzing motion patterns on the sensor, the BMI563 reduces the latency and power overhead associated with continuously streaming raw data to the application processor.
The BMI563 provides a comprehensive interface set: primary MIPI I3C®, I²C, and SPI, along with two interrupt pins (for I²C/I3C) or one interrupt pin in 3-wire SPI mode. A secondary AUX interface is available for connecting an OIS (optical image stabilization) actuator or another peripheral. The platform’s timing specifications—latency below 0.5 ms, time increment of approximately 0.6 µs, and timing resolution of 1 ns—are particularly relevant for XR and robotics, where sensor data must be time-aligned with camera frames or control loop iterations. This temporal precision enables tighter sensor fusion and more stable tracking.
The BMI563 is engineered for three primary application clusters: inertial SLAM in autonomous robots, high-dynamic motion tracking in XR controllers, and motion-based automatic scene tagging in action cameras. In robotics, the combination of extended range and vibration robustness supports the accurate pose estimation that SLAM algorithms require, even when the robot is moving aggressively or operating on uneven terrain. In XR, the saturation-free gyroscope data preserves controller orientation through fast movements, preventing the tracking loss that can break immersion. In action cameras, the sensor enables automatic tagging of scenes based on motion characteristics—identifying a jump, a descent, or a rapid pan—without post-processing.
The BMI563 is currently available as samples to direct customers, with high-volume production planned from Q3 2026. Designers integrating the sensor should account for the 8 KB FIFO, which allows batched data reads and reduces bus contention in systems with multiple sensors. The wide supply voltage range (VDD 1.71–3.6 V, VDDIO 1.08–3.6 V) provides flexibility in power rail design. For robotics applications, careful attention to mounting and mechanical decoupling is essential to realize the sensor’s vibration rejection capabilities; the package should be rigidly attached to the structure whose motion is being measured, with isolation from sources of high-frequency mechanical noise where appropriate.
The BMI563 is defined by its willingness to operate where motion is extreme and precision cannot degrade. Its extended full-scale range prevents saturation, its vibration robustness preserves signal integrity, and its Edge-AI engine adds a layer of on-sensor intelligence that benefits both robotics and XR. For engineers building systems that must track fast, forceful motion with unwavering accuracy, the BMI563 offers a sensor architecture aligned with the demands of the application rather than adapted from a lower-dynamic design.