Arduino Step Tracker

Chung Shun Ko, Leon Ge

What to achieve

Challenges

• With Arduino to create a compact and low power consumption device to count step
• Implement smoothing and filtering algorithm
• Implement peak detection algorithm
• Able to reset the count without restarting whole system

​

• We designed our own 3d printed case
• We also had a lot of challenges dealing with data types during the implementation.
• Fit the data and improve accuracy under constrained resource such as small RAM and processing speed

Technical knowledge

• We applied the in-class knowledge of step detection by calculating magnitude
• Since it's almost impossible to run scipy and other libraries, we needed a different algorithm for smoothing and detection.
• We implemented an simplified version of the Z-score smoothing algorithm.
• By using a z-score peak detection, we can detect steps in real time.

Implementation

- We used an accelerometer similar to a phone to detect leg movements as well as forward acceleration.

- Using a moving average array to do smoothing. A z-score peak detection will allow for peak detection based on previous values instead of hard coding.

- Based on the z-score plot, we can see that when the is a z-score above 2-2.5 or so, there is a step taken.

- Using the z-score method can accommodate for different intensity of acceleration, instead of hard-coding a magnitude threshold.

​

Problem Statement

Potential usage of the project

• Although we have learned to do step detection with pre-recorded data and potentially using phone data, we decided to do a little challenge.
• We wanted to see if it's possible to do real-time data processing on a small arduino microcontroller to make a step tracker.

• There may not be a lot of practical application of this project because there are way better step detection algorithms out there, including phones and wearables.
• This project was done as a learning experience and for fun, and it's also always cool to make hardware stuff.

Summary

• Result
• We are able to implement a variation of the z-score smoothing algorithm to combine data smoothing and peak detection.
• The device is functional and counting steps.
• Future improvements
• Dynamically adjust z-score threshold for variation in steps.
• Improve the smoothing and filtering algorithm to better process accelerometer data, including “fake” steps.
• Also include activity recognition

​

​

References

• https://stackoverflow.com/questions/22583391/peak-signal-detection-in-realtime-timeseries-data/
• https://chrisruppel.com/blog/arduino-analog-signal-input-smoothing/
• https://dganesan.github.io/mhealth-course/chapter2-steps/ch2-stepcounter.html
• https://gcc.gnu.org/wiki/avr-gcc
• https://hackaday.io/project/167533-faster-speed-how-optimize-math-and-process-tasks/log/168869-faster-sqrt-but-less-accurate