Fitness trackers and training computers put to the test03/14/2016
Sports and computer scientists of the University of Würzburg have taken a close look at current fitness trackers, activity sensors and training computers. They wanted to find out how accurate these devices are and whether they are suitable for athletes and everyday users to manage their training. A first study has been published recently.
Sportwissenschaftler und Informatiker der Uni Würzburg wollten herausfinden, wie akkurat aktuelle Fitnesstracker und Trainingscomputer messen. Und ob eine effiziente Trainingssteuerung für Athleten und Alltagsanwender damit überhaupt möglich ist. Auf dem Bild: Eine Laufuhr des Herstellers Garmin. (Foto: Marco Bosch)
Athletes modify their training every day. They do this to boost their performance and also to prevent injury. To be able to adjust the scope and the intensity of training, athletes need an overview of their performance data such as heart rate, maximum oxygen intake and other parameters. The data are recorded and stored by portable mini-computers or sensors, the so-called wearables.
These digital helpers are becoming increasingly popular in recreational sports, too: Recent surveys put the global market at around five and a half billion euros. “This is clearly the trend in the entire sports and leisure industry across all sports and applications,” says Professor Billy Sperlich and he adds: “But a lot of the devices have not been evaluated yet.”
The sports scientist leads the department of integrative and experimental training science at the University of Würzburg. For the current study, he and his colleague Peter Düking carefully scrutinized several products.
Estimation of energy expenditure inaccurate
The scientists did not perform a mere benchmark test as currently published in various runner's and sports magazines. Rather, they designed the study to determine what the individual devices are capable of and which of the measured parameters are important to manage training efficiently in the first place. "The devices available in the market today are already capable of measuring many biomarkers, but they are not entirely accurate," says Sperlich.
An example from recreational sports: Many people want to lose weight. Fitness trackers can help them get an overview of how many calories they use on one day. But: "Some devices have significantly underestimated the real energy expenditure when doing sports," Sperlich says.
With his team, Sperlich wants to compile as much data as possible to track down patterns. To assure data privacy and facilitate preparation and interpretation, Sperlich and Düking have teamed up with Andreas Hotho, Professor of Computer Science. "We hope that the huge volumes of data (keyword: big data) will enable us to determine the real effect of training on an individual more efficiently and to detect variables that we wouldn't have deemed relevant for performance before," Sperlich further.
Wrist-based heart rate monitors often lack accuracy
To manage the training in an optimal way, different sensors have to be combined. The gadgets examined by Sperlich, Düking and Hotho in a first step use different techniques: The classic chest strap provided good measuring accuracy, whereas the optical wrist-based monitoring of the heart rate is still in its infancy: "The measurement results of optical wrist sensors were frequently inaccurate, especially during intensive physical training," Sperlich says. This method uses LEDs that send light waves into the skin through tissue and blood vessels. The light is either absorbed, transmitted or reflected in this process. A lens located between the LEDs uses the reflected light, which is different depending on the levels of blood flow in each cardiac cycle, to derive the pulse.
Moreover, other sensors fitted into clothing or mattresses can provide information about an athlete's quality of sleep, measure the level of dehydration using skin sensors, monitor the core temperature with a kind of ear plug and determine the blood circulation in muscles, the maximum oxygen intake, and of course, the body weight.
Most commercially available devices are limited to duration, distance, speed and difference in altitude combined with heart rate and quality of sleep.
The next generation will be capable of much more
The electronic assistants of the next generation will also take other factors into account such as the runner's general well-being. "Unfortunately, the devices are not yet capable of sensing when somebody has an infection and should therefore train lightly in the morning," Sperlich names one example and adds: "But this will come, manufacturers are on the right track."
According to Sperlich, trainers and doctors do not have to feel that their existence is threatened by the smart gadgets and health monitors in the future. Some training computers are suitable when it comes to planning heart rate based training and achieving the set goals. Wearables also benefit people who find it more difficult to assess their physical capabilities, for example, having been injured or sick for some time.
Nevertheless, the two professions mentioned above will continue to be irreplaceable in everyday life. The scientist believes that it is reasonable and important to keep to personalized training schedules and to regularly consult your physician who has access to all health-related data. "Moreover, how we feel is the most sensitive stress marker we have," Sperlich says.
If you don't feel well, you should leave your running shoes in the wardrobe and stretch or get some rest instead.
"Comparison of non-invasive individual monitoring of the training and health of athletes with commercially available wearable technologies" by Peter Düking, Andreas Hotho, Franz K. Fuss, Hans-Christer Holmberg, and Billy Sperlich. Published in: Frontiers in physiology: journal.frontiersin.org/article/10.3389/fphys.2016.00071/abstract