#LoveIrishResearch BLOG: Optimising exercise with wearable motion sensors and your smart phone

Martin O’Reilly is an IRC Enterprise Partnership Scheme Postgraduate Scholar based at University College, Dublin and working in partnership with Shimmer and Realtime Technologies Ltd., He is also member of Ireland’s Olympic Handball team, which is aiming to qualify for the Euro 2020 Championships.

Many people undertake resistance training programmes to gain improvements in health, mood, strength, athletic performance, weight loss and quality of life. However, due to cost and availability issues the majority of gym-users train without the expert supervision of personal trainers. This means novice exercisers receive minimal feedback on their training. Without expert feedback, adherence to exercise programmes diminishes and the risk of exercising with poor technique increases greatly. This results in slower progress towards training goals and a heightened risk of injury.

In my research, I investigate if wearable motion sensors, such as those in smartphones, can effectively analyse exercise technique and serve as an input to a feedback based mobile app. I hope that this will help people exercise more safely and effectively. One of my key research goals is to achieve high quality exercise analysis with a minimal wearable sensor set. This makes end user systems more cost-effective and practical for users. Through my research, I have developed a system whereby a single wearable sensor can detect what exercise somebody is doing and how many repetitions of it they are doing with greater than 98% accuracy. A single sensor can also tell if somebody is exercising with proper or aberrant form. For instance, acceptable and aberrant squat technique can be detected with over 95% accuracy using a single sensor on the shin.

I love my research because it allows me to combine my two main passions; sports and problem solving. It is very exciting to see that when you apply the right algorithms to data from inexpensive, widely available sensors you can learn very interesting things about the way somebody is exercising. What exercise are they doing? How many times are they doing it? Are they doing it at the right pace? Are they moving safely and with biomechanical efficiency? What can they do differently to improve their technique?

I am now incredibly eager to investigate whether giving gym-users answers to all the above questions can ultimately improve their exercising experience and outcomes.  By wearing just a single wearable motion sensor and using my exercise tracking smartphone app, I hope that people will adhere better to their exercise programmes and avoid exercising with aberrant technique. This, in turn, should help them reach their training goals more efficiently whilst also avoiding injury along the way. Hopefully, in the near future my research will be an inexpensive widely available tool to guide everybody through their exercise programmes.