Exercise Motion Analysis AI
We are very excited to announce a brand new feature that we have been working on called fitsw.ai (pronounced Fitsway). This new feature uses exercise motion analysis AI (Artificial Intelligence) to analyze exercise movement from a variety of aspects. Then, the system produces a movement score based on these aspects. It is a great way to provide an objective report to your clients based on their movement. You as the trainer are likely able to analyze your client’s exercise movement on your own. However, our tool can also pick up imperceptible deviations as well. It then gives you a report that you can use as a talking point to communicate what you saw with your client.
We have demonstrated this product at a few conferences and the trainers that have seen it have been very enthusiastic about it. Not only can it give the client and trainer insights into movements, but it also can serve as a motivation tool for clients.
How? Well, we as trainers know often takes months for a client to see results on the scale. Even if a client has been really working hard for a few weeks it can be discouraging to not see that scale move right away. However, even in the first few weeks of training, a client will likely make drastic improvements in their movement mechanics. FitSW.ai can show that progress to clients right away. We have seen clients receive analysis scores 30 points higher after 2 weeks of training. This can serve as motivation to clients to know that they are getting stronger and moving better even if the scale hasn’t changed yet.
This data-driven personal training can be quite powerful. Click the image below for a demonstration.
How To Use Exercise Motion Analysis AI
Ok, so it’s easy to see the benefits of fitsw.ai but how do you use it. That’s easy too. All you need is a computer or phone with a camera. That’s all you need. Setup your computer or phone and login to the fitsw.com website.
In the side navigation you will now see a button that says “fitsw.ai Beta”. When you click this, your device will activate your camera. You can click the “How To” button for directions but we will also list them here:
- Press the green “Start” button & wait till it turns into a red “Stop” button (the first time always takes a few seconds)
- Step back until your full body is in view of the camera and make sure to face the camera straight on
- When ready, place your hands behind your head and a 3 second countdown will appear
- When you see GO! do 5 squats
- View your results and have fun trying to get a better score!
- If the Restart button is not in view scroll down
As you can see, its super simple to have the system analyze your clients squats.
What Does the Results Analysis Data Mean
Ok it was simple to do the anaylsis, but what does the results analysis data mean. Well there is a lot here so lets break it down:
The Top Graph: Rep Tempo
The top graph shows your up and down movement over time which is essentially your rep tempo. The green dots above it are where you are starting each rep and the red/orange dots are where the rep was finished. So you can see that there were five reps and all were fairly consitent in terms of tempo in the above image.
On the first 2 reps you can see a lot of red on that graph. That red indicates that at that time there was center line deviation. That means that the center of the persons body went off-center at those times. This data feeds into metric scores which we discuss below.
Next Graph: Left Knee Right Knee Graphs
The next graph shows left knee and right knee movement over time. Again it also shows 5 cycles – one for each rep that was done. The red dots on the above graph are points where the knee moved inward over the threshold.
The 5 Rep Center Graphs
These 5 graphs show the left/right deviation of the body’s centerline as it went down and up for each rep. So for rep 5 on the way down (blue dots), the person deviated slightly to the right. Then on the way back up (green dots), the user deviated back to the left.
The goal with these graphs would be to have a line of dots straight up and down without deviation either way. The deviations measured are very small amounts so some is ok, but the less the better.
The data from the graphs above, all feed in the scores output in the Metric Scores section. The higher the score, the better. Going down the list:
- Rep Tempo – this score is based on how consistent the tempo of all 5 reps was. You want to make them as consistent as possible.
- Depth Score – measures that you went deep enough in the squat.
- Rest Tempo – at the top of each rep there is a slight pause. This score measures the consistency of that pause.
- Center Line – scored based on how much the center of your chest deviated from the center while doing the 5 reps.
- Left Knee – scored based on how consistent your left knee movements were with both tempo and movement to the left or right.
- Right Knee – the same as the above but for your right knee.
- Final Score – A total score based on all of the other scores.
As you can see there is quite a bit of data that is quickly analyzed to give you and your clients an objective score. Please let us know if there are any other type of aspects you feel that it would be worth analyzing.
Right now the fitsw.ai analyzes squats but we are also tuning the AI for several other movements including balance tests, FMS movements, jump analysis, and more. If there are any particular movements you are interested in, please let us know at email@example.com and we will prioritize them. These new exercise motion analysis ai movements will be ready soon but please give us any feedback that you have in the meantime.