AI Splits Fitness From Trainers
Does that mean future gyms would have a supervisor who doesn’t train? Maybe so!
Imagine a fitness trainer in your gym who provides a personalized approach to everything from diet to training schedules and then supervises the routine with a behavior that is guided by one’s own goals and abilities. And what if this ubiquitous fitness trainer and instructor cannot be seen but only heard as one works out a sweat or cools down in the confines of the gym?
This is artificial intelligence (AI) now going beyond smart mirrors, training apps and cameras, and taking over the actual workout routine once you get to the gym. This is no more science fiction as Lumin Fitness located at Las Colinas in Texas now boasts of instructors who can be heard but cannot be seen – because they do not exist in the real world.
A report published by MIT Technology Review says the gym has several virtual trainers such as the energetic young coach Chloe, the disciplined Rex or the more mellow Emma and Ethan who are warm and friendly but quietly tough as well. Of course, all this doesn’t mean there’s no human present in case you need physical help. But, this supervisor needn’t be a trainer.
How does the AI-led gym work?
So, how does the whole thing work? The gym uses wall-to-wall LED screens and motion tracking sensors to guide the clients through their workout routine. Then there’s also a set of algorithms that provide insights using a variety of data points that the system captures during the process as well as after it.
Of course, the basics are all there. So, one can pick the preferred voice and coaching style right upfront. The next step is the gamified fitness experience where the LED screens and algorithms provide an enhanced experience that encourages complete movements and achievement of specific fitness goals based on one’s body capabilities.
As one becomes a regular visitor, the integrated sensors within Lumin Fitness take over to track the number of repetitions and form during workouts. In fact it monitors each movement of the member and if done incorrectly, does not count it as a completed repetition. This avoids potential injuries besides tracking progress over time, with recommendations for increasing weights or trying different movements or reducing repetitions.
The power of instant workout analytics
The gym, which opened barely a month ago, has seen good traction with people ready to try out the new way at the cost of just $25 for seven days. The studio can accommodate up to 14 people at a time who can participate in functional training classes comprising squats or sit-ups or take up workout using gym equipment.
The workouts are done facing the LED screens at a designated station that not only serve as mirrors but also as sensors that track the person going through a workout as well as the specially created equipment such as dumbbells, medicine balls, skipping ropes etc. The owners say that machine-learning models have already improved the algorithms.
The gym has also come up with its own app (available on both Android and iOS) for selecting the coach, the workout schedule etc. Of course, one needs to carry around earpods in order to listen to the coach’s instruction with one’s own choice of music then playing in the background to keep one company.
There’s still a need for human presence
The MIT Review quotes Lumin Fitness cofounder Brandon Bean to suggest that the gym does have a fitness professional who supervises the routines, but going forward they needn’t be trainers. He notes that it’s more about having someone around in case something goes wrong as otherwise the AI trainer controls the whole process of regimen, feedback and motivation.
In fact, the LED screens also help during warm-ups and cool-downs with a humanoid figure showing the motions as a visual aid. However, once the workout begins, the screens shift to some motivational games. For e.g. you could fill up a virtual basket with balls by completing a sit-up done in the correct way.
What’s going down well with the visitors is the ready analysis that is available of their workouts as well as the exact quantum of improvement over time. The company says it also plans to train the system to recognize times when people find the workouts too easy and the AI can recommend heavier weights or different movements.
Of course, the success of this experiment eventually depends on the personality type of a fitness enthusiast. There are those who love interacting with machines and there are others who feel that human presence is the only true motivator.
