4 ms·

I have a Basis B1 band that I've been using on and off since launch. The one thing it is really poor at is capturing heart rate information and other metrics du

I have a Basis B1 band that I've been using on and off since launch. The one thing it is really poor at is capturing heart rate information and other metrics during periods of heavy activity. If you are working out or performing an intense activity, the device frequently dismisses the activity as noise for one and stops capturing the data. e.g. when my heart's pounding after a heavy lift or a sprint and is ready to pop out, the Basis band either stops recording or shows my heart rate that's no where close to true. I've exchanged emails with them in the past about their algorithms dismissing such activity as noise but they've maintained that they cannot discuss the algorithms they use to calculate the metrics and it's not a device to be used for such purposes; it's more about giving you a rough estimate as you carry on your day to day life. As far as sudden heart rate increases during the night, you could possibly co-relate it to your dreaming; I woke up sweaty once after having a nightmare and my skin temperature & heart rate #s were definitely skewed by a good margin as compared to a normal night of sleep.

mojuba 13y ago
Judging from the LED they have on the underside they are using optical oximetry: basically you can see the blood flow dynamic based solely on the levels of oxygen under the skin. That explains why it's not accurate when you move: once you move your arm your blood flow becomes so unstable that it's practically impossible to "see" the pulse there. This is a known problem.
- desigooner 13y ago
  True. They do use other sensors to gauge the skin temperature and such. Coupling the data from that plus the accelerometer, you would at least be able to improve upon the metrics calculated. From my understanding, it's not so much of the "seeing" of the pulse being the issue as the algorithms they use to cut off the data collection, identifying it as "noise". As a result, there is errant or missing data collection during rest periods (when optical oximetry wouldn't suffer due to the movement). If there's a sustained movement, using data such as skin temp and accelerometer makes it easier to identify what's noise and what not. Their calorie tracking really suffers due to the errant tracking of heart rate and the intense activity periods; comparing my off-days with workout days, the difference in calories burnt is a meagre 100-150 calories. A moderately brisk walk alone burns about 200 calories an hour. I suppose I was expecting some basic learning algorithms that adjust the data collection according to your day to day activities; but then this is the first iteration of the product, so there's hope. At this point of time, for me, it is just a sleep tracker + pedometer + a basic overview of the day.

mojuba 13y ago

Judging from the LED they have on the underside they are using optical oximetry: basically you can see the blood flow dynamic based solely on the levels of oxygen under the skin. That explains why it's not accurate when you move: once you move your arm your blood flow becomes so unstable that it's practically impossible to "see" the pulse there. This is a known problem.

desigooner 13y ago
True. They do use other sensors to gauge the skin temperature and such. Coupling the data from that plus the accelerometer, you would at least be able to improve upon the metrics calculated. From my understanding, it's not so much of the "seeing" of the pulse being the issue as the algorithms they use to cut off the data collection, identifying it as "noise". As a result, there is errant or missing data collection during rest periods (when optical oximetry wouldn't suffer due to the movement). If there's a sustained movement, using data such as skin temp and accelerometer makes it easier to identify what's noise and what not. Their calorie tracking really suffers due to the errant tracking of heart rate and the intense activity periods; comparing my off-days with workout days, the difference in calories burnt is a meagre 100-150 calories. A moderately brisk walk alone burns about 200 calories an hour. I suppose I was expecting some basic learning algorithms that adjust the data collection according to your day to day activities; but then this is the first iteration of the product, so there's hope. At this point of time, for me, it is just a sleep tracker + pedometer + a basic overview of the day.