By studying more than 600,000 data point per week, Microsoft is hoping to find ways to keep athletes healthier.
Read about the latest sports tech news, innovations, ideas and products that impact players, fans and the sports industry overall at SportTechie.com.
A gold medal in the Olympics can come down to milliseconds. So it is important to be able to prepare for the games as much as possible without the obstacles of injuries and illnesses taking away precious training time.
After teaming up with the Australian Institute of Sport (AIS), Microsoft and BizData, Australia’s Olympic team may have found a way to pinpoint when an athlete might suffer injuries or illnesses. By using emerging technology and creative data analysis to track the day-to-day training of athletes, the coaches are able to manage peak performances and predict the drawbacks. It is reported that the athletes could lose up to 20% of their training time because of being sick or injured, and thus be unable to meet their goals.
“We used predictive analytics and machine learning to better understand the relationship between training loads and injury and illness. We now know that an athlete needs to maintain a high to moderate chronic training load,” said Nick Brown, deputy director for performance science and innovation at the AIS, in an interview with Microsoft. “We capture a whole lot of information from 2,000 athletes each week—including about 300 data points per athlete, or if you like, 600,000 data points a week.”
The collected data is stored every night through an Azure SQL Database to the Athlete Management System. Then during the next morning, the custom algorithms are updated to let the coaches modify each training routine for the day according to the received data.
With these changes, the organization was forced to change the way it manages and handles data, which was previously spread across multiple hard drives. The new, more organized way has said to have helped them manage data in a much more efficient way.
The project is still in its pilot phase but has already shown to be helpful by being able to predict which athletes have a high chance of being injured within the next three days.