Dr Lyndell Bruce, Sport Scientist at Deakin University: “Harnessing data is the biggest opportunity in sports tech right now”

With the Paris Olympics just around the corner, we can expect a bunch of records to tumble. While 99% of that success will be down to the efforts of athletes and their coaches, that vital 1% – the crucial difference in many cases – will be due to data and sports science. And people like Dr Lyndell Bruce, Sport Scientist at Deakin University, who we’re delighted to interview here.

As well as being an accredited sport scientist, Dr Bruce is the Associate Head of School (Partnerships) within Deakin’s School of Exercise and Nutrition Sciences, the world’s number one sport science school. Her research explores expert performance and factors contributing to expertise and the development of expertise, and that’s something Australia has in abundance.

In this interview, Dr Bruce reveals some of the latest trends in sports science, including developments at Deakin University. Such as a decision-making tool for coaches to use in matches. “One of the most exciting aspects of this system is that it can bring together the strengths of AI alongside the strengths of a coach or analyst to improve the outcomes for team performance,” she told us.

“System users can view the probability of their team winning the match based on a team’s current in-match performance but also see what performance indicators they should change to increase their winning probability.”

And this is just one example. Read on to discover how digital twins are being used by coaches, the impact computer vision technologies are having, where esports fit into Dr Bruce’s research and why the way athletes and coaches capture and harness data is so important to their success.

Related reading: Let the games begin: Paris Olympics put AI to the test

Can you give an example of a complex problem in sports that you have been involved in tackling with technology?

Coaches in professional team sports are often provided with reams of data during a match that they need to align with their first-hand perceptions of what’s occurring. Coaches may have a gut instinct on what is happening but want to use the data to support their thoughts.

Often, it’s not possible to organise and analyse the data in a succinct time efficient way to meet the coach’s time constraints. Changes often need to be made in a matter of minutes otherwise the opposition may gain momentum and swing the game in their favour.

Mitch Aarons, a PhD student at Deakin University, designed and created a decision support system (DSS) for coaches to use to assist in their match-day decision-making. More information can be found here.

The DSS is an AI-based solution built upon thousands of matches from previous seasons. He explored various models to find the one that was able to best predict match outcomes and then used this model to build the system around.

The DSS itself has some customisable features that enable individual teams to focus on key performance indicators that are most important to them, which might be based on their game style. System users can view the probability of their team winning the match based on a team’s current in-match performance but also see what performance indicators they should change to increase their winning probability.

One of the most exciting aspects of this system is that it can bring together the strengths of AI alongside the strengths of a coach or analyst to improve the outcomes for team performance.

What are some examples of AI being used in sports that stand out to you?

There are a few that provide the greatest excitement and potential to have an impact on performance. They are the concept of a digital athlete or twin, the use of computer vision for performance analysis and the potential for development of technical and tactical aspects of performance.

The digital twin enables sports scientists to input what we know about an athlete, such as their physical capabilities, movement patterns, injury risk, performance, training history and health data into data simulation models. We can then apply stimuli, such as changes to training loads and technique changes to see the impact this could have on the athlete.

This could provide insights around injury risk and performance optimisation that are currently not possible in the timeframes that can be achieved through simulations. Future applications may include a digital team being used to test game strategy against different opponent structures to optimise team tactics to increase the potential of team success.

With a background in skill acquisition and performance analysis, computer vision technologies are exciting as they have the potential to expedite the performance analysis process and provide opportunities to explore game strategy optimisation. The ability to combine movement data alongside performance data (ie, technical and tactical) has the potential to revolutionise the way we approach performance analysis and athlete tactical development.

In all these developments and potential applications, at the forefront of the decision-making process and considerations need to be the ethical application and confidentiality of the athletes.

Recommended: Marilou McFarlane, CEO of Women in Sports Tech: “It’s rewarding beyond measure to see how technology has impacted the growth of women’s sports”

Do you consider esports as part of your remit? If so, why?

As sports scientists, we apply our knowledge and skills across many different domains of human performance, including sport, emergency services, first responders, military contexts and esports. There is often a philosophical debate around applying sport science principles to combat (ie, shoot-fire) esports competitors, but this is one that individuals need to assess and decide upon supporting or not.

The health and well-being of esport athletes need to be considered as they often spend large amounts of time in a sedentary position (there are exemptions such as Zwift of course). Despite their sedentary nature, physical and mental fitness play an important role in an athlete’s ability to maintain the concentration required to compete at a high level.

Applying a skill acquisition lens to esports is very interesting as we can start to explore through gaze behaviours what the expert players are looking for on their screens versus players who are not as skilled. We can also look at how individuals and teams use and provide feedback, both in competition and post-competition, to explore avenues for improvement.

Obviously, these athletes spend a lot of time training, so we can look to maximise their training environments to improve performance.

How do you stay up to date on the latest technology developments in sports?

As a sports science researcher, there are a number of different ways to stay connected to technology developments. These include attending conferences, reading the latest research publications, social media and mailing lists. However, one of the best avenues, which can be achieved through some of the aforementioned ways, is through networking and regular communication with colleagues. Nothing beats having a conversation (virtual or otherwise) with a colleague and exploring what they are currently involved in.

Working at a university often means you are able to hear about new technology while it’s still in development and you can begin to think of the application, in my case, for sport. A bigger opportunity lies in the potential collaborations within a university. We can work with our colleagues to explain problems or inefficiencies and then work together to create innovative solutions.

I subscribe (via email) to some newsletters that provide updates on technology or sport research (eg, Australian Sports Technologies Network, Catapult and university mailing lists). There isn’t one specific platform or people that I follow on social media but try to cast a wide net and check in regularly on these platforms.

Related reading: Kristof De Mey, Co-Founder of OnTracx: “It’s often forgotten that implementing something new always comes with challenges, resistance and setbacks.”

What do you believe is the biggest opportunity in sports tech right now?

Harnessing the data. Our capacity to collect data from a vast array of available has never been higher and in the foreseeable future will only increase. We need better processes and systems to harness this data and distil it into meaningful and useful information. Within performance departments of sports clubs, data comes from multiple sources and varying formats. Opportunities lie in creating user-friendly data management solutions to manage this data. The integration of the different data sources to then create user-friendly outputs continues to be a source of frustration for performance staff.

A further opportunity lies in being able to improve efficiency in any aspect of an organisation. Opportunities exist to automate processes to free up performance staff to engage in higher-order thinking, have conversations with athletes, innovate, and engage more with coaches.

Irrespective of the technology developed, sporting organisations are still going to need the best people to facilitate the interpretation of the data and translate data output into meaningful solutions for coaches and athletes. Humans are needed to further distil the information into smaller bite-sized pieces for coaches and athletes. Transferable skills such as people management and communication skills will be more critical than ever to ensure coaches and athletes buy into the outcomes generated from the data.

When it comes to technology, what are the most common mistakes you see sports organisations make?

One of the biggest mistakes organisations can make is collecting data but not having the personnel to use that data in a meaningful way. It then sits, often in an unusable format, in a folder on some shared drive (if it is even extracted from the device). If athletes (and coaches) are not able to see how their data is being used or that it is providing value to them, they will decrease their compliance in providing the data, or falsify information. Athletes need continual education about what data is being collected and why. The level of education and regularity will vary depending on each individual athlete, gender and experience.

Another issue is the rush to keep up with the Joneses. Clubs may purchase the latest technology, promote it on their social platforms and think that this will attract players, support staff, sponsors, and fans. However, we often see clubs being lured in by marketing private providers trying to sell products which have not been externally assessed for their validity and reliability. Often these tools or ‘new toys’ will lack ecological validity. They might look good and even be fun to play with, but there is often limited or no transfer to performance. Worse still, they may detract from a player’s performance. The ability to be able to scientifically assess a new tool or piece of technology is critical for sporting organisations and those who work in them.