Last Updated on December 5, 2022 by pg@petergamma.org
- We made some preliminary tests on a treadmill at a speed of 1 km/h for one hour and a Garmin Fenix 6 without foot pod. We found at this speed, that the Fenix 6 is very inaccurate. The inaccuracy strongly depends on the speed we moved our arms.
- We also made some tests with the Android app a training tracker from Dr. Ing. Rainer Blind. The app is not available anymore in the Google Play Store, but as a GITHUB open source project:
https://github.com/rainerblind/aTrainingTracker
- But we can still find the app as an AKP download in the web. We made some tests with a training tracker and a Suunto pod. At around a speed of 4 km/h, the app accurately counts steps from the foot pod. We did not do any comparison tests to other step counters. But we tested it for a while for walking mediation, and we never observed, that a training tracker missed any steps.
- A foot pod has the advantage, that it counts also steps when we do not move our wrist, for instance if we are wearing some bags.
- Be aware that for instance Garmin watches with the protocol indoor walking which counts steps does not count the steps of the foot pod, but only the steps from the wrist. At least in older Garmin watches we tested.
- When we paired a training tracker to a Stryd Fod pood, we observed, that a training tracker is also capable of counting steps below 4 km/h, which is great for walking meditation.
- A training tracker only counts steps from ANT+ foot pods, but not from BLE foot pods. This is because of the BLE protocol does not support step count, according to Rainer Blind. At least a some years ago.
- We also found out that a training tracker has issues when using different sensors. We observed several times, that suddenly pairing to an ANT+ foot pod was not possible anymore, for instance after using BLE sensors. The only way we could solve this problem was a complete hardware reset of our smartphone. In our case it was a Samsung Galaxy S5 with built in ANT+ support.
- A downside of a training tracker step counting is, that we do not know of any software platform which can process a training tracker step counting. The steps are stored in a .csv file. They can for instance be imported into Microsoft Access for further processing.Another downside of this that we have on this way many .csv files in our Access Database which need further processing. One for steps, one for instance from a Garmin watch with converted .fit file. Data analysis on that path is difficult.
- A great alternative are Apple watches. According to this instruction:
- Steps from Apple watches can be stored into InfluxDB and vizualized in Grafana, which is geat. But we did not test this yet. And we do not know whether Apple watches supports activity steps, or only step counting during the day.
- We also asked Rob ter Horst to do a test with a training tracker and the Stryd foot pod, since we suppose that this combination is eventually one of the most accurate step counter which is available, unfortunately without success to this date. We do not plan to publish any accuracy tests.
- There are also other options available. The Adafruit Open Source Bike Computer Pyloton which supports speed and cadence, can be modified to support foot pods. The same applies to Reto Roellins Rasperri PI ANT+ and to Balena Health with the Polar H10 chest strap. But these are options for users with advanced coding skills, or at least for users, who are ready to spend a lot of time to implement these features. But according to our own little experience, there is a lot of potential to improve the implementations for step counting for practicing walking meditation. With Python, InfluxDB and Grafana, we are convinced that these kind of features will be supported much better sooner or later.