How big data analytics played a part in Vienna INEOS Challenge
Congratulations are in order for Eliud Kipchoge and his team for achieving a feat that no human has achieved before, it was indeed a demonstration of grit, will power and big data!
In this narrative, we highlight the technological advances that fueled the performance we witnessed. We review the evolution of methods for extracting kinematic information from athletes, observing how technology has progressed from laborious manual approaches to optoelectronic marker-based systems – the motion analysis systems which are currently most widely used in sports biomechanics that collect kinematic data which later are exposed to potent mathematical algorithms otherwise commonly referred to as Machine Learning with a view to optimize and predict future performance.
In a detailed documentary on National Geographic titled Breaking 2, we are given a preview of a project by Nike to break the two-hour barrier for the marathon. Nike announced the project in November 2016 and organized a team of three elite runners who trained for a private race.
The event was held on the Formula One Autodromo Nazionale Monza race track in Italy on May 6, 2017. Suffice to say the historic event witnessed on Saturday the 12th was a culmination of years of meticulous planning, data collection and analysis. Its not a coincidence that Monza formula one track was chosen. Historically, the formula one racing strategy is fueled by data analytics and there are troves of data that are mined by the analyst to make key decision points. However, despite the similarities in approach, extracting and analyzing data from mechanical objects such as racing cars is different from extracting the same from human movements.
In its rudimentary form, athletes have been using data to measure performance since the invention of the stop watch. The study of human movement within sports biomechanics has made considerable progress over recent decades. For instance gait is one of well-identified biometrics that has been broadly applied for human identification at a distance based on their motion style. Due to its wide applications, researchers have developed various technologies to monitor and to analyze human movement in different activities, such as running, walking, jumping, football, golf, etc. For the past few years, human motion capture has been accomplished using optical motion capture system and magnetic tracking system.
Data capture and Analysis is transforming the athletics world by modelling behavior and logging performance. This is driven by the rapid growth of technology, communication system and transmission through telemetry, miniature body-mounted sensors have been considered as alternative methods to monitor and to collect the kinematics parameters of the human movement athletes. These sensors include accelerometer, gyroscope, goniometer, magnetometer, and their combinations which gather vast quantities of information and combined with GPS systems and video analysis can measure upto 1,000 data points per second.
This wearable sensors record data such as speed, power, temperature, ground contact, cadence and vertical oscillation. Information on previous competitors and athletes can also be mined by way of comparison. This creates a reservoir of data on one individual athlete and how they are performing. All this this raw data is then fed into analytics engines to produce bespoke reports. This is then used to forming a training regime tailored to an individual athlete, and inform post competition performance analysis.
Athletics coaches are bringing in tech innovators to give them this data edge, with the latest discrete wearable technology able to capture every professional conceivable metric. The Ineos challenge for instance was powered by technology infused into Nike shoes that Kipchoge wore, a future edition of Nike’s Next% marathon shoe.
The shoe is a result of athletes, sport scientists, engineers and designers closely collaborating throughout the entire process of design, testing and manufacturing and data was at the heart of it. A company called Stryd has produced a system that enables miniaturization of data capture by attaching a devise to running shoes which emits data to high powered analytics software’s to produce digestible results. Data analytics was also used to pre-emptively deal with potential injuries by measuring asymmetries in movement thus reducing the chances of critical injuries occurring during his training preparation.
The choice of track was also driven by data, a flat surface track with few bends ensured consistent speed not to mention the V shaped formation that was to shield him from the wind and building a slipstream for him through aerodynamics meaning he conserved his energies.
Advancements in computing power, sensor developments and machine learning will yield more improved algorithms able to convert data into more sophisticated insights to help athletes maximize their potential.
Timothy Oriedo, Founder and CEO – Predictive Analytics Lab