How NEXUS solved the equation: From burpees to machine learning engines
With the launch of the NEXUS platform (mobile app and PUSH wearable device), we have solved one of the most significant problems in the functional fitness space: The ability to measure your workout, quantify specific aspects of your training session, and provide actionable metrics to improve your performance.
To achieve this, the biggest hurdle for us to overcome was that the athlete could not be expected to interact with any device or technology during a metcon-style workout or WOD. Once started, it is a race against the clock, so the athlete needs to “set and forget” any measurement device. Step one to solving this problem was finding the right tool for the job. The PUSH Band is a wearable accelerometer/gyroscope that has been on the market for five years having proven itself by cutting its teeth in pro sport and collegiate weight room training environments, where it remains the dominant wearable device in the space. Given the varied nature of a WOD, which can include gymnastics, medicine balls, assault bikes, skipping, etc., a wearable device was a mandatory requirement, as none of the other devices on the market were able to measure all of these exercise modes seamlessly. For these reasons, the PUSH band was the device of choice to solve the problem of quantifying the sport of fitness.
Problem Solving Using AI
With the right hardware in place, step two to solving this problem was to utilize the power of artificial intelligence to build upon the current PUSH algorithm engine. The workflow of PUSH is a set-to-set workflow, where the user or coach starts and stops the device at the beginning and end of each set. Since this wouldn’t work for an athlete completing a WOD, to get around this, we developed two algorithm engines utilizing machine learning and artificial intelligence, one for exercise detection and the other for automatic set transitions. For those who have never heard of machine learning, it is a subset of artificial intelligence that uses the power of computers to “learn” to find answers and solve problems using large datasets in order to train the engine.
The first algorithm can look at the signals coming from the band and characterize them based on unique features or “fingerprints” and can determine what exercise is being performed. From this, the algorithm can also determine if the signal is a true exercise movement, or if it is a “rest period” such as drinking water or walking around. Layered on top of this is the second algorithm, which can determine when the athlete switches from one exercise to the next or when completing a set or round and moving to the next. Combining these two algorithms, we are now able to measure your WOD from start to finish, requiring zero user-interaction with the technology for the duration of the workout.
NEXUS vs Heart Rate?
The NEXUS metrics are considered “external” metrics that represent the movement produced and loading sustained by the individual. These result from the forces created when interacting with the weight room equipment and environment. Heart rate (HR), which is one of the most commonly utilized metrics in the exercise field and one of the only metrics currently used in the functional fitness space, is considered an “internal” metric. Being an internal metric means it is the body’s response by the cardiovascular system stemming from the movement produced and load sustained during the WOD. While HR has been relatively easy to measure, it only tells part of the story. With NEXUS, we are finally able to tell the whole story by being the first to measure and report key metrics from your workout performance.
Since the metrics that NEXUS can measure are new to the functional fitness space, it is beneficial to understand what we can measure, how we measure it, and why it matters to you. Keep your eye out for our next posts as we will dive deeper into the metrics that NEXUS provides and how they can help you get the most out of your training and performance.