Validation study of positioning accuracy of smart watches under multiple outdoors application scenarios

An increasing number of elderly people are adopting smart watches as pedometers to track their routes during outdoor running. However, there is a lack of proof regarding the positioning accuracy of these devices. Contrary to cellphones or vehicle positioning systems, which update location information at a certain frequency, smart watches often lack longitude and latitude data unless the user is in motion. Hence the objective of this study is to propose a method for assessing the positioning accuracy of smart watches with these characteristics. This study utilized an RTK surveyor to generate reference values, a Coros Pace2 smart watch to provide testing data (referred as watch data), as well as a high-speed camera for correcting positioning data. Altogether, 3 tests with 660 watch data were conducted in an outdoor urban environment in Beijing. The results show that average position errors are 15m in single point positioning error test, 6.91m in path positioning error test, and 4.53m in dynamic path positioning error test. The positioning errors in the latter tests are lower than the manufacturer's claim of 10m. This research validates the effectiveness of the testing method and illustrates the importance of pre-market testing for smart watches to ensure accurate positioning.