Wearable sensor technologies hold the key to unlocking novel performance assessment and there is interest in the development of sports technology. There are alternatives to video cameras as a means of data capture and with the advances and miniaturization of sensors, these can now often be unobtrusively attached to the body which record data directly and can be processed in ‘ real time’ to present information to the coach. Whilst these provide substantial functionality for analysis, using them requires a considerable investment of time, especially for a coach or researcher analyzing a swimmer over multiple lengths.
There is a plethora of software available (e.g., Kinovea ®, Quintic ® and Dartfish ®) to aid in the analysis of these videos, some of which are adapted to mobile devices. An underwater camera can be added to a trolley to move with the swimmer along the pool which would capture a larger dataset of the underwater stroke. Similarly, there is an increasing use of mobile-based technology (e.g., iPad ®), filming from the side of the pool, above the water limiting the view of underwater actions. This can lead to misleading conclusions being drawn on a small sample. However, there are limitations to the use of 2D methods within swimming, for example, the camera(s) are generally fixed to the pool capturing a limited number of strokes. Standard (2D) video technology has become easily accessible due to a general reduction in price, leading to regular use within coaching. Particularly for coaches, 3D camera systems present with limitations of considerable setup time and cost, which limits their general accessibility. Scientific analysis in swimming typically involves either 2D or 3D camera systems. The digital systems data interactions with various other team members.
The use of technology in sport can help facilitate this. This shows that the coach will have access to the data from each of the science disciplines to be able to make a decision based on the periodic cycle of the athlete/team, but also, that each scientist has some understanding of their counterparts’ data in order to create a ‘fuller’ picture before making recommendations to the coach.
The flow of data for this can be seen in Figure 1, developed from Hughes. This will allow the practitioner to discuss data with the coach and create extrinsic feedback to help the athlete. Ideally, there will be a coaching team around the coach and athlete/team, which is often only seen at the higher levels of sport. To overcome this bias, coaches can work with sports science practitioners to employ scientific methods to enhance the level of detail they can use in their coaching practice. Previous research has shown this to be bias when considering gross movements, so when considering detailed biomechanical factors, there will inevitably be a greater degree of inaccuracy. This is equally true when coaching swimming, where coaches will provide extrinsic feedback, involving demonstrations and verbal instructions/descriptions, based on the premise of what the coach could see. Most swimming research tends to focus on descriptive stroke characteristics, such as stroke rate, because they are more ‘ readily observable’. This customization requirement demonstrates that single based devices will not be able to determine these phases of the stroke with sufficient accuracy. The developed algorithm was developed using a search window relative to the body roll (peak/trough). Four swimmers required individual adaptation to the stroke phase calculation method. The calculated factors were compared to the same data derived to video data showing strong positive results for all factors. Twelve swimmers, equipped with these devices on the body, performed fatiguing trials. Using multiple units on the body, an algorithm was developed to calculate the phases of the stroke based on the relative position of the body roll. The purpose of this study is to use a multiple inertial measurement units to calculate Lap Time, Velocity, Stroke Count, Stroke Duration, Stroke Rate and Phases of the Stroke (Entry, Pull, Push, Recovery) in front crawl swimming. Objective data on swimming performance is needed to meet the demands of the swimming coach and athlete.