Tech Trends in Fitness
Originally when I set out to create this, I was going to build a resource that outlines the different pieces of tech and also talk about how even the non-electrical machines you use in the gym are tech. As I have mentioned in other posts, I am not a computer genius, but I do excel in tech if we broaden the scope to gym equipment, as I have built several bespoke pieces, some of which do things that are not available on the market and I am very proud of that. The breakdown and comparison part of equipment has plenty of resources that are done very well so doing so would not provide anything new. I want to break down some of the negatives with a lot of the emerging tech, specifically electronics, in the fitness industry. I believe my background in behavior change science, statistics and in business may provide a unique and worthwhile perspective into what is actually going on around the who, what, where, why of these devices and if you should actually buy them. By you, I mean specifically the individual you because some of these will be context dependent so I won't be making a sweeping yes or no.
Outline of discussion
Most AI is not personalized
It can't see you move
If it can, it does not correlate that data with your subjective inputs (pump/ connection/ pain)
Fitness trackers are expensive habit trackers and trendy status symbols
Not good for long term use
Distracts from main goal
You start focusing on the measurements not the why it is being measured in the first place
Not supported by evidence to improve long term habit adherence
More features does not mean better or even that it actually tracks the thing it says it does.
How do they actually work
What do they actually track
Geolocation-
Movement-
Body temp-
Vibration-
Changes in blood volume-
How do they track
Even direct measures like heart rate and steps have interference affects from other movements but many devices do this quite well of an r-value of around .75-.95 for just a general vibes check where you have to get more subjective and self comparative is in your recovery and total effort levels experienced
(this is the Ai overview from google that does a decent breakdown that I wanted to walk you through the context of it) (Everything in parentheses is my explanation)
Steps:
Accelerometer: Detects movement in all three axes (up-down, side-to-side, front-back).
Algorithm: Interprets accelerometer data to identify and count steps. (generally more direct and less reliant on population norms)
(More limited by accuracy of the devices monitor itself)
(so if the device reports high accuracy than you are probably fine)
Distance:
GPS: Tracks location and calculates distance traveled.
Accelerometer: Estimates distance based on movement patterns.
(same here if the device reports high accuracy than you are probably fine)
Heart Rate:
Optical heart rate sensor: Shines light into the wrist and measures blood flow through the skin.
Algorithm: Converts blood flow data into heart rate. (has more to interpret so accuracy is import, so if you actually read your result and do more intense exercise, watch) https://www.youtube.com/c/TheQuantifiedScientist
Calories Burned:
Accelerometer: Tracks movement intensity.
Heart rate: Provides information about effort level.
Algorithm: Estimates calorie expenditure based on movement and heart rate data. (makes an assumption based on norms, in general these are not very consistent because organ size plays such a large role in BMR as well as small movements in NEAT they may not be tracked
Sleep:
Accelerometer: Detects periods of inactivity.
Heart rate: Provides additional information about sleep quality.
Algorithm: Determines sleep duration and quality. (makes an assumption based on norms and correlations)
(I would recommend something accurate, obviously, but probably ignore most of the specific data. Many people get excited about more numbers and feel more confident in having all these specifics, most track the same thing, others just make extra guesses that get further and further from direct measurements. Total time asleep and how much you moved around and how much your hr changed.)
Other Metrics:
Some fitness trackers may also track additional metrics such as stress levels, breathing rate, and body temperature. (stress is where
Accuracy:
It's important to note that fitness trackers are not always accurate. Factors that can affect accuracy include:
Device quality
Movement type (e.g., swimming, biking)
Wearer's body size and weight
Environmental conditions (e.g., temperature, humidity)
(and your experience, have fun being your own case study and determine what the values mean for you as you get used to the device you picked, if you don't want to do that at all just know it is probably not going to be exact in anything it measures, some more so than others)
Overall:
Fitness trackers can provide valuable insights into physical activity levels and help motivate users to improve their health. However, it's essential to use them with caution and to understand their limitations
Why do they track that
They use these measurements listed in the “what” section and make A LOT of inferences. Which is fine as long as you know the actual accuracy and what is actually being tracked. My last post went into details about businesses providing the public with there own data, well here we are again.
Things it says it tracks
Sleep
Menstrual cycle
recovery
Where are you supposed to get from using this
The idea is that this builds a habit and having the physical device reminds you and having a log inspires you to compete against yourself and in some cases a community
Things no one is talking about
These devices do not actually know how well you sleep or your cycle or your recovery or strain percentage or your total “body battery” they are using equations developed from population norms and different covariates like height weight sex activity level will change what equation it may use. You may be two standard deviations from the mean of the bell curve, meaning every thing has a variance and these are complex things being tracked that are challenging to study perfectly in a lab setting, your fitness ring does not know down to the decimal point what your score is. Here is why that is actually totally fine. What you should actually use these things for is self comparison. Maybe you are only 75% recovered but you feel totally fine and you are making progress, don't cut your workout because a stupid watch had a yellow color on it. Use it to figure out how much you can push and recover and then once you find that point use whatever the watch says as your point of reference. Maybe your resting HR is lower than normal so it says you are more recovered than you actually are. Maybe you read online that taking a nap will recharge your body battery but then it reduces your sleep efficiency for the week and you think it is because of something else.
Cadence lock- When your trackers start to misinterpret the rhythm of your cardio for your HR which comes back to being more specific with what the device is actually tracking. This is part of why chest strap monitors are more accurate. They are closer to your heart and can more easily distinguish between your heart beating and the movement and vibrations and shifts that come with exercise
Who are you (discussion on data usage)
In the terms and services you agree to you are very likely giving them access to and the ability to the sell the data that the company collects for you yet they don’t do a great job of admitting their own limitations of what they are tracking on you this piece feels very unilateral and seems like companies want it both ways.
What is next?
What I would like to see these companies do and I myself have been working on, is a way to have a more complete picture so it is more individualized and less affected by inter-individual variability. Big tech companies love ecosystems but I have yet to see something that tracks the visual data you would get from a tonal and compares it to the data tracked by a wearable and better yet the data you input yourself. This will be where things trend especially for those more invested or competitive maybe in athletic facilities and of course some fanatics. An ecosystem that that stores this data an uses it in comparison to itself can help improve the accuracy on all sides.
Data needs context