FITBIT INSPIRE 2 Testing

INTRODUCTION

Dr. Rob is a postdoctoral scientist based in Vienna, Austria. He performed a complete set of scientific tests in Fitbit inspire 2.First, he tested the quality of sleep tracking against the EEG monitor. Second, he checked the heart rate and finally checked the accuracy of the step counter. It was very interesting to see that the cheap inspire 2 performs equally well as always. He likes to provide backgroun information that is very important to inspire 2. Fitbit inspires 2 is one of the cheapest suitability Fitbit has on its list in his opinion. It has the most important features provided by Fitbit: sleep tracking, 24-hour heart rate measurement, and step counting. In addition, it has a longer battery life of up to 10 days compared to other Fitbit products, considering the purchase of the Fitbit inspire 2. The Fitbit Inspire 2 has no onboard GPS, no spo2 sensor, and no Spotify control.

TEST

In the tests, he focused on three of the main features of the Fitbit inspire 2: the sleep tracker, the heart rate monitor, and the step counter. He started with sleep tracking for the sleep comparison. He wore the Fitbit-inspired 2 to bed for 3 nights at the same time. He also wore this portable scientific EEG device and he recorded himself using an infrared camera.The EEG machine records both muscle and brain activity. Several of his colleagues are using it in research investigations; it is known as the hypnotise z-max. He carefully scored each hour of the night for the various stages of sleep while listening to the EEG recording. In addition, he also exported the data from the Fitbit inspire 2 and converted it to a format. He could use the infrared recording; he could check what his movements were like and see if they inspire 2 correctly predicted when he was awakened. First, he had a look at the accuracy over the three individual nights after which he would do a statistical overview analysis there.

TEST OF SLEEP STAGE

The first night, he recorded on top as they were measured using the EEG device on the horizontal axis. They had the time of night he went to bed quite late a little bit after midnight on the vertical axis you have the different sleep stages that are deep sleep light sleep rem sleep and awake the sleep stages are plotted in the order that is also usually displayed in research a similar plot but now for the sleep stages as they were recorded using the Fitbit inspire 2. He marked in purple, there's a pretty decent match between the inspire 2 and the EEG device for the second deep sleep segment, the exact position was slightly off but overall the rough positioning and amount of deep sleep were correct, compared to other sleep trackers the inspire 2, seems to do well based on this first night most deep sleep would generally occur at the beginning of the night.

THE INSPIRE 2 PERFORMANCE IN SLEEP CYCLES

However many sleep trackers greatly overestimate the amount of deep sleep at the end of the night and the inspire 2 does not seem to have this problem now of course he needed to look at the other two knights as well to see if the Fitbit inspired 2 performed consistently well but first let's look at rem sleep which was marked there in red. He saw the very impressive accuracy of the Fitbit charge 2. It almost exactly captured the ram stages, it only missed. This last ram segment though would likely have been a relatively lighter part of his sleep. Already, considering that it was close to the end of the night and he woke up naturally. This day overall the first night bodes well for the Fitbit inspire 2. To see the sleep cycles he added non-rem sleep in blue and again marker rem sleep in red. Each sleep cycle starts with a combination of deep sleep and light sleep together called non-rem and always ends in rem again non-rem is marked in blue and rem in red based on just the Fitbit data. He would be able to pick up on the sleep cycles. He went through, except for the end of this last cycle overall the inspire 2 performed very well when it came to sleep cycles.

TIME CHECKING

Next, looking at the times, he was awake marked in green. The inspire 2 performed very well, picking up on both moments of awakening when it came to sleep, also called sleep onset and the end of sleep. For this night, the inspire 2 correctly predicted the moment that he fell asleep but he woke up slightly sooner than he did. The next night, he tracked to start again with deep sleep similar to before deep sleep was marked in purple. The inspire 2 correctly picked up on the deep sleep he had. This was again very impressive, it did pick up on some extra deep sleep at the end of the night. That wasn't there, but these were just small segments. Overall, it still performed much better than most of the competitors. Next, sleep marked in red again. They saw a very good match, it did. However, miss the first short rem sleep segment at the beginning of the night. This also means most of the sleep cycles except for the first one which would have been hidden due to the missing rem sleep segment.

However, he said that he was still impressed looking at the awake moments. These were detected well. There's a small extra segment of awake time. However, he might already have been in a lighter sleep by

that time. Since he did have a significant awake moment right after when they looked at sleep start. There was a slight delay in the Fitbit detecting sleep. Though this was minor and the sleep end was detected quite well.

Finally, looking at the last night, again they saw good deep sleep detecting the ram stages, and therefore, the sleep cycles were detected almost perfectly again. This was very impressive. The inspire 2 detected a bit more awake time than he had. Though, before sleep start and sleep end were pretty

much spot on. He said,'' These results looked very promising.” However, to get an even more objective view of the results, they calculated some statistics regarding the consistency between the Fitbit inspire 2 and the EEG device.

SLEEP TRACKER TESTING

First, they looked at the total percentage of each sleep stage that the EEG device on the left and the inspire 2 on the right predicted overall. This agreement was quite good, they saw that the inspire 2 predicted a little bit too much deep sleep which was consistent with what they saw in the individual nights. It also predicted a little bit too much awake time which is also what they saw on the individual nights. The inspire 2 predicted too little light sleep the total amount of rem sleep. However, it was spot on predicting exactly 19.1. Although, even more important than these total percentages was checking. If the inspire 2 predicted the correct sleep stages at the right time. That's what he had displayed. They had the sleep stages according to the EEG device and on the left the sleep stages according to the inspire 2. Now, each column, sums to 100 meaning that they could see what percentage of each of the actual sleep stages was recorded, as each sleep stage by the inspire 2.

They first looked at deep sleep they had seen that two-thirds of what was deep sleep was also predicted as deep sleep and the remainder was predicted as light sleep. Now, this was pretty good since both of these were part of non-rem sleep.

Next, looking at light sleep they saw that most of what was light sleep was also predicted as light sleep. If light sleep was confused most of the time with deep sleep, they also saw that most of what was rem sleep were also predicted as rem sleep. This again strengthens the conclusion that the inspire 2 can in most cases detect the different sleep cycles. If there was any confusion rem sleep was confused with light sleep but never with deep sleep.

Finally, by far most awake time was also detected as awake time and if there was any confusion this was with light sleep which made sense. Since this was the stage closest to being awake overall the statistical results strengthened his feelings about the sleep tracking quality of the Fitbit inspire 2. They seemed to be of the same quality as the Fitbit charge series, which were among the best fitness wearables for sleep tracking; he liked to test the inspire 2 for many more nights, maybe around 10 to 40 nights total to see. He was really happy with the sleep tracking accuracy of the Fitbit inspires 2.

HEART RATE ACCURACY TESTING

For the next set of tests, they had a look at the heart rate accuracy of the inspire 2. to test the heart rate accuracy. He compared it to the polar h10 ECG chest strap which was generally considered to be one of the most accurate consumer devices available for heart rate measurements. He wore both the inspire 2 and the polar h10 chest strap for 3 spinning sessions and 1 weightlifting session. That way he could check his heart rate at different heart rate ranges, and they looked at those results. There, he displayed an overview of the heart rate accuracy. Each dot there had a single heart rate measurement with on the horizontal axis the value according to the poder h10 chest strap and on the vertical axis, the value according to the Fitbit inspire 2. The blue line indicates perfect agreement so any measurement along this line had roughly the same value for the polar h10 and the inspire 2. The red line indicates those

measurements where the value according to the inspire 2 was half of the actual value according to the polar h10. The reason he added that was because in the past he had seen that many devices measure half the actual heart rate when they made a mistake the more measurements there were in a certain area the darker blacker colour.

COMPARISON OF THE INSPIRE 2 AND ECG

Overall, there was a pretty good agreement between the inspire 2 and the ECG chess trap most points were along the blue line only there in the lower heart rate range. They saw significant deviations between the inspire 2 and the polar h10 since the points were below the blue line this meant that the inspire 2 detected a too low heart rate in these moments during the individual training sessions to see if they could find the underlying cause of the first spinning session along the horizontal axis they had the time and his heart rate was along the vertical axis in blue. He plotted his heart rate according to the polar h10 chest strap and in red was his rate according to the inspire 2.

SPINNING SESSION

For this, first spinning session, they saw pretty good agreement the red line almost perfectly overlaps the blue line. He took three short breaks in the spinning session where his heart rate would dip interestingly. At the end of two training segments, there was a short spike in his heart rate that was not there right. During these moments, his heart rate dropped quite suddenly and he suspected that the algorithm of the inspire 2 had some difficulty for a brief moment dealing with that drop in heart rate looking at the next spinning session. They saw almost identical results. There was almost perfect agreement with a small spike at the end of one of the segments. This third training session showed a few more problems. Again generally, the agreement was pretty good. It was at the end of the segments that coincide with a sudden drop in heart rate where the inspire 2 struggled to keep up with this decrease. Finally, they looked at a weightlifting session. There they could see what caused the problems. They saw in the overview plot in the beginning that weight lifting causes sudden spikes in heart rate and at the same time the muscles in his arms were tense. He suspected this combination. was what made it difficult for most optical heart rate sensors to correctly detect his heart rate Unfortunately, the inspire 2 is no exception; it follows the general patterns in his heart rate but misses the spikes in heart rate.

Overall, the Inspire 2 performed well when it came to heart rate measurements, it had a few small artefacts when he was doing cardio workouts but these were just small issues during weight lifting it did have some more serious issues so he wouldn't buy it for that specific purpose.

However, this was true for almost all wrist-worn wearables. The Fitbit Inspire 2 also features a step counter to see if this counts his steps accurately. He went out and took exactly 4 000 steps in segments of 1000 steps to get an accurate step count. He manually counted his steps using this telecenter.

The results as he mentioned for the step counting test he went out and took 4 times exactly 1000 steps he wore the Fitbit inspired 2 on his left arm which was also what it was set to in the settings of the app. He alternated holding the step counter in his left and right hand for each set of 1000 steps which was what the right and left labels refer to right.

These were the steps counted by the inspire 2. There was only a small deviation from the actual 1000 steps. He took each time in each case the inspire 2 overcounted by between 14 and 63 steps which was not a lot but might still indicate a consistent pattern of slightly counting this matched the data he had shared previously for other Fitbit devices. If they compared the inspire 2 to 2 other devices he wore at the same time the amaze fit band 5 and the Mi Band 5. The inspire 2 is the only device that systematically overcounts steps. Though it was still very close to the actual 1000 steps each time out of these 3 devices. It was the one most deviating from the correct number of steps. Overall though, he still thought it performed well enough.

THE COMPARISON OF FITBIT CHARGE 4 AND

FITBIT INSPIRE 2

In that test, he also wanted to test if the inspire 2 gives any false positive steps that he meant, does it count any steps? When it's not supposed to count steps for instance when typing or cycling. He also wanted to see how it performs on other people couldn't say anything yet about false positive steps but so far results seem to indicate that the inspire 2 is pretty good at counting steps when it's supposed to count steps. Overall, he was very satisfied with the accuracy of the inspire 2. Sleep tracking is especially among the best he had seen and seems to be on par with the Fitbit charge series; it's particularly good at detecting rem sleep and deep sleep. Though it predicted a little bit too much deep sleep also. It's sometimes slightly too sensitive at detecting awake moments when it comes to heart rate measurements. He was quite pleasantly surprised.

He'd noted some issues with the heart rate measurements of the Fitbit charge 4 before and the inspire 2 seemed to have fewer of these issues.  Now, he needed to analyse more days of exercise to be sure but overall it was pretty accurate. The only thing it struggled with was weight lifting  but this was a problem that all wrist-worn heart rate monitors have shown in his tests. So far, when checking the step counting, he found the inspire 2 to be pretty good. It counted the correct number of steps while walking. However, he found Fitbit to be notorious for counting steps when they're not supposed to count steps, for instance when he was cycling or on his stationary bike.

CONCLUSION

So, he recommends buying the Fitbit inspire 2. If we're interested in great sleep tracking and good heart rate and activity tracking at a reasonable price. The inspire 2 might be the best option for us. Though, the list price is around $ 99. It is on sale for prices between 70 and 80 dollars. However, if you want a built-in GPS and an oxygen saturation sensor then the more expensive Fitbit charge 4 might be a better option for us. Finally, he should mention some of the limitations of the data that he showed here. First of all, he just tested the Fitbit inspire 2 on him and it'll be interesting to see how it performs on others second he only tested it for a limited amount of days.

WITH SPECIAL THANKS TO DR. ROB