Data Sheet

October 2017 BNO080 Datasheet 1000-3927
www.hillcrestlabs.com © 2017 Hillcrest Laboratories, Inc. All rights reserved. 33 / 57
Prediction
Time (s)
Alpha
Beta
Gamma
Max
error
(deg)
rms error
(deg)
Noise
StdDev
(deg)
0.028
0.3030725439091
0.1132958963849
0.0027762197131
1.96°
0.19°
0.019°
0.005
0.3643672983581
0.0931655637068
0.0025557744608
0.25°
0.01°
0.003°
0.01
0.3222569517214
0.0945467654507
0.0022830845312
0.54°
0.03°
0.005°
0.02
0.2905555274225
0.1018361851286
0.0023953721024
1.26°
0.1°
0.011°
0.03
0.3084584610521
0.1160861868509
0.0028497585549
2.14°
0.22°
0.021°
0.04
0.3396316005051
0.1237361273333
0.0030844488660
3.23°
0.4°
0.034°
0.05
0.3548447957314
0.1224432975403
0.0032236047505
4.8°
0.64°
0.051°
0.06
0.4091677424529
0.1250943899234
0.0035648974825
7.03°
0.93°
0.072°
Figure 2-2: HMD mounted head motion prediction
For other applications contact Hillcrest Labs.
2.3 Environmental Sensors
The BNO080 provides support for environmental sensors connected over an I
2
C interface (separate to the host
interface). Currently the BNO080 supports:
• Bosch Sensortec BME280 pressure/humidity/temperature sensor
• Bosch Sensortec BMP280 pressure/temperature sensor (only BMP280 or BME280 should be populated)
• Capella Microsystems CM36686 proximity and ambient light sensor
If the sensors are not required for a particular application the I2C bus should be correctly terminated with pullup
resistors as the BNO080 attempts to discover the sensors at reset. Proper termination will ensure correct
behavior of the BNO080.
The ALS and proximity sensors typically require calibration as these sensors are typically placed behind glass or
plastic. Contact Hillcrest Labs for details.
2.4 Classification System
With the proliferation of sensors available within mobile devices, there is an increasing interest in providing value
by understanding the context of the device or device holder. Classifying the context based on the sensors
available is an active area of research with an emphasis of the classifiers being ‘always on’, so low power is an
inherent requirement. BNO080 supports a number of classifiers that generate events upon detection of a
particular context or motion.
2.4.1 Stability Detection and Classification
Analysis of the motion sensors allows BNO080 to classify stability. BNO080 provides two virtual sensors to
quantify stability: a stability detector and a stability classifier.
The stability classifier uses both the accelerometer and gyroscope to distinguish between three levels of stability:
• On table: the device is likely on a table or other stable surface
• Stable: the device is likely being held, but held in a stationary manner
• Motion: the device or device holder is in motion
The stability detector uses the accelerometer to distinguish between stability and motion.
Both the classifier and detector have configurable thresholds for the levels of stability. These thresholds are
stored as FRS records as described in the SH-2 Reference Manual [1].
The stability detector has an acceleration threshold which is the aggregate acceleration of the device. The default
value is 0.784m/s
2
. The stability detector also includes a time threshold. The acceleration of the device must stay
below the threshold set for the duration of the time record to register as stable. The default value is 500ms.