As I have specified in the following blog post: https://ciprianaa30.wixsite.com/aurismartwatch/post/issues-reading-sensor-data-nicla-sense-me
The Nicla Sense ME uses the Bosch Sensortec AI Suite for a supervised machine learning system implementation, where the user can create as many entries, data sets and models as they want, with all the sensors on board to perform specific outputs in specific scenarios.
The program and SDK/API can be found at the following link: https://www.bosch-sensortec.com/software-tools/software/bme688-software/
As part of the features offered by the Nicla Sense ME, it includes an already trained supervised machine learning model on the IMU data (gyroscope, accelerometer + time information) in order to detect the type of activity the user is performing.
Such as:
Standing
Walking
Running
Cycling
Driving
The issue encountered was generated by incorrect/incomplete documentation available here: https://docs.arduino.cc/tutorials/nicla-sense-me/cheat-sheet#activity
The reference error is:
Serial.println(String("Activity info: ")+ activity.toString());In order to get the activity output we need to use activity.toString());.
The issue is that the syntax is incorrect, after a bit of back and forth with the Arduino Support Team, they found out that the actual syntax to use is:
Serial.println(String("Activity info: ")+ activity.getActivity());
The toString method is used in every other sensor output from the Nicla Board ME, but the Activity function uses its own different call.
Here is a demonstration of the sensor in action:
Here is the Arduino IDE test file:
Comments