The Era of AI-Based Fitness

Original article was published by Aidlab on Artificial Intelligence on Medium


A few months ago, Huami introduced its next-generation artificial intelligence toting chipset — the Huangshan 2, that the company says allows any activity tracker to learn your behavior to improve your health and performance. In 2019, Aidlab released its engine that can match motion models in real-time by detecting data with the motion sensor and heart rate sensor on Aidlab’s wearables, and then intelligently determine the exercise mode of a user. At present, Aidlab can automatically detect 16 exercise modes, including walking, running, cycling, push-ups, or squats, which cover 95% of daily exercise scenarios of users. Users do not need to perform tedious manual operations, but can enjoy a senseless selection of smart exercise modes.

Exploring the Possibilities of AI in Everyday Activities

Let’s start by establishing for whom AI would work the least. An overall active person with a well-established set of healthy habits might find little use in an AI activity tracker. The reason is simple: the healthy routines have been in place for years and need no monitoring or correction. However, even those with ingrained healthy habits can still be making mistakes that are difficult to spot, precisely because the pattern of unhealthy routines has been repeated over and over again. In that aspect, having the AI assistant can be useful in further refining good habits.

As said, the more you wear such wearable, the more it learns about you and your routines improving. This lets the algorithm pick out recurring patterns that can be responsible for a variety of lurking ailments, accuracy, or bouts of fatigue. Here are some of the activities the AI-equipped assistant can measure and analyze today:

Sleep patterns — sleep is crucial to maintaining good health. It’s necessary for proper recovery, memory consolidation, brain waste disposal, and plenty of other things.

HRV (heart rate variability)measuring heart rate variability helps you manage stress and schedule your workouts.

Respiration rate — tracking your breathing pattern during a workout is a good indicator of training effort. If you’re jogging, you should have only a slightly elevated respiration rate. If it’s higher, it means you’re going too hard at a workout meant to be easy.

Movement analysis — it will tell you if you’re moving too little and sitting too much. Daily, weekly, and even monthly analysis can give you actionable insight.

Posture — if you have to sit for a long time, at least make sure you’re doing it right.

Heart-rate trends — following your monthly resting heart rate fluctuations will help you track your training progress (that is if you’ve taken up a training regimen and are consistent). HR trends can also be used to measure stress or an underlying illness.

So who can benefit from AI-based activity trackers?

It is obvious to see that the technology will be used by most professional sports teams as they practice, to track biometrics, workout levels, diet, sleep patterns and overall fitness. Some wearables can also be used during games for post-game performance analysis. AI systems can help detect and identify signs that are indicative of athletes developing cardiovascular or musculoskeletal conditions. Other examples are:

People with Little to Moderate Physical Activity Who Want to Change. The sheer amount of data gathered from a week-long measurement can be enough of an incentive to implement healthy habits, or at the very least recognize that there are areas in your current lifestyle that need to be addressed to mitigate the possibility of various risks revealed by the analysis.

Recovering Patients. With a hefty set of guidelines provided by a doctor, it’s not always easy to follow every item on the list through. The AI assistant will not only help you maintain good habits but the data gathered from weeks-long measurement can also serve as an excellent foundation to gauge your progress during recovery. This is especially valid for devices with advanced heart-rate monitors.

Busy People. Small but crucial habits like drinking enough water, stretching, or walking more tend to suffer the most in the tight agenda of busy folks. The AI engine will collect data on your daily behavior and identify patterns that need your intervention.

Athletes. With an activity tracker that measures HRV, you can learn about your body’s readiness for more training. HRV can be used to determine if you’re recovered enough to absorb another hard training stimulus or whether you should take it easy. Remember, exercising with elevated stress levels and high workout load won’t expedite positive adaptations but further compound fatigue.

Risks?

Buzz surrounding AI and its application in fitness have been pretty spectacular and as experts say, it will only increase. Health is no different from other areas where artificial intelligence is actively used — where there’s an opportunity to implement efficiency in everyday activities, AI reveals patterns that need to be changed to improve operations. Insights drawn by AI activity trackers can help eliminate bad habits and introduce healthy routines.

We are also warned of the dangers AI technology can introduce. The most obvious one is that AI systems will sometimes be wrong and that injury or other health-care problems may result. But is AI for health, wellness and fitness really dangerous and does it deserve the hype that’s generated around it?

We will find out in the upcoming years.