AI in sport:
The key lies in combining digital innovation with human expertise

Sports scientist Andreas Kramer on the potential and limitations of artificial intelligence

Andreas Kramer, a sports scientist specialising in sports medicine at Hellersen, can well understand this trend. “AI is a great source of inspiration. It can create detailed training plans, tailor and plan sporting activities, and also provide a targeted recovery analysis. AI is fundamentally changing the way people train today,” explains Andreas Kramer. But he must also clearly point out the limitations of AI and where risks lie.

In grassroots sport, smart apps are taking on the role of a digital coach. Based on performance metrics, current fitness levels or even sleep and stress indicators, they create personalised training programmes, analyse progress and provide motivational feedback. “This can be a valuable source of support, particularly for athletes without direct access to professional coaching,” the sports scientist continues.

The importance of AI is also growing in elite sport – particularly when it comes to analysing large sets of data. Systems capture GPS, video and biometric signals to optimise training processes, load distribution and tactical decisions. Wearables, including smartwatches, with integrated AI provide precise real-time analyses of parameters such as heart rate, changes in performance or recovery status. “On this basis, training stimuli can be controlled in a more targeted manner and overloading can be avoided at an early stage,” explains Andreas Kramer.

“Nevertheless, artificial intelligence cannot replace human expertise. Experience, empathy, intuition and situational judgement remain central elements of successful training support,” he adds. Another important aspect is the prior physical assessment of the athlete and, in particular, taking pre-existing conditions into account. In such cases, consultation with an experienced exercise scientist or a doctor is essential.

Furthermore, potential sources of error in the AI’s interpretation of data, as well as an excessive reliance on automated recommendations, must also be taken into account. For instance, an AI-based analysis may lead to incorrect training guidelines – particularly if exhaustion or overexertion are not recognised as such. Wearable manufacturers very often provide no information on how data is processed or how recommendations are generated. “It’s a bit like driving on autopilot. Would you rely on it completely when it comes to your health?”

Andreas Kramer sums it up: “AI can be a valuable addition to sport. The future of training lies in the combination of digital precision and human expertise and intuition.”