Sports Analytics

At a glance

In interdisciplinary teams made up of sports scientists, engineers and data scientists, Fraunhofer IIS carries out research and development projects on ways of combining sport and technology. The primary focus is on the swift detection of situation-relevant events and the preparation of this information for the user. This is made possible by the use of intelligent algorithmics that filter events out of the positioning data, prompting a quick response with suitable measures.
We decided to pool expertise in this area because of our many years of experience in the processing of sensor data in sports and the use of state-of-the-art technologies.

AI and Machine Learning

The terms “artificial intelligence” and “machine learning” are omnipresent these days. Fraunhofer IIS uses these technical innovations to add value. The basic idea behind these methods involves a computer generating relevant analyses out of existing data. Thanks to the input of humans (chiefly sports scientists in this case) and to highly advanced training methods for learning algorithms, even complex performance indicators can be realized.
To be able to use the latest research results from this field, Fraunhofer IIS collaborates closely with the Friedrich-Alexander University Erlangen-Nürnberg (FAU).

Tiny Machine Learning

Tiny Machine Learning (TinyML) is a research area in the field of machine learning and describes the optimization and execution of AI-based processing chains on embedded systems.

In industry projects, we offer our customers automation solutions for AI development for ultra-low power applications or consult data science teams on optimized workflow and reliable testing in seminars and workshops.

Official match data in pro soccer

In the first and second divisions of Germany’s professional soccer league, video-based tracking systems are currently being used to capture the ball and players by means of machine vision (tracking rate: 25 Hz). In addition, events such as passes, tackles and shots on goal are captured manually by commercial match data providers.
Fraunhofer IIS processes this data into player- and team-related performance indicators in order to enable high-quality analyses. Examples of this include metrics on passing networks and the continuous evaluation of the compactness of a team.

Tracking systems, inertial measurement units and video integration

Both, when developing its own localization systems for sportspeople (tracking rate of up to 3000 Hz) and when integrating and analyzing commercial systems, Fraunhofer IIS is able to draw on over 10 years of experience. By combining the technology with inertial measurement units (IMUs), existing tracking systems can become even more precise or supply even more added value through orientation data. The combination of video and tracking data also delivers new insights into large swaths of sporting activity. For example, the entertainment offered to viewers can reach a new level by displaying live information on the TV screen.


Sports science

As a technology-oriented research institute, our success is also based on the way we incorporate the classical sciences in order to use well-founded knowledge in projects. Accordingly, alongside engineers and computer scientists, our sports scientists are a fixture in sports-related projects, where they play a decisive role.


Process mining

A relatively new area of research, process mining is concerned with the analysis of processes. In the context of sports, these methods (process discovery, conformance checking, and process enhancement) can be used to analyze and interactively design training exercises and even generate tactical models of whole matches. In terms of practical applications, this field of research came to prominence through automatic analyses of medical procedures, which means that established methods already exist.



New methods of machine learning and artificial intelligence, combined with precise real-time tracking, allow a new depth of game and training analysis.



Automated analysis methods in sports provide technically, tactically or physiologically important parameters for added value in sport applications.



Automated action recognition in combination with machine learning and hierarchical event recognition allows the automatic generation of game statistics.


Sensors in soccer

Real-time analysis of soccer players' performance and the course of play using cost-effective sensors and machine learning to detect speeds, actions and events for trainers and players.


Validating wearables in basketball

The validation of commercial wearables enables optimal training and game control of athletes in the NBA (National Basketball Association).

Sensor technology for boxing

Real-time measurements of the acceleration curve and speed of punches using specially developed miniaturized sensor technology in combination with sensor-based artificial intelligence.

S3Engine: Sports Scene Search Engine

Fast sports-scene search engine. Based on deep learning, our S3Engine technology enables fast searches for similar game scenes in sports, evaluates them and suggests solutions.

We offer...

Evaluation and advice for sports tracking systems

In the world of sports, users have a broad range of tracking systems at their disposal. Depending on the intended use, the various technologies differ with respect to quality and suitability. To be able to evaluate their quality in detail, our experts have focused on carrying out comprehensive system evaluations with a high degree of automation and standardization. We have the capability to test both indoor and outdoor systems and calculate sports-relevant evaluation parameters (e.g. running distances), whereby the precision of the available reference systems extend down into the submillimeter range. In addition, there is the option not only of evaluating the characteristic values of movements themselves, but also of their effects at the physiological level, such as heart rate and breathing rate, which are often recorded in parallel with the movements.


Development of algorithms and performance of studies for technologies in sports

Do you need R&D support to develop your algorithms for sporting analyses? Then Fraunhofer IIS is exactly the right place for you. We offer a unique testing environment (L.I.N.K. test hall) for professional tests and comprehensive studies. Fraunhofer develops and tests algorithms for applications such as orientation estimation, training load monitoring, the recording of KPIs (key performance indicators), pulse and HRV measurements and even interactive training exercises and technical/tactical analyses in team ball sports. Our portfolio is aimed at the manufacturers of sports tracking systems (radio-based, GPS, video), of combined systems including vital parameters and particularly the manufacturers of wearables.


Analysis of official match data

Match analysis is a fixed component of performance diagnostics for a wide variety of professional sports. It is used to evaluate match situations in order to record the strengths and weaknesses of sportspeople. In this field, the whole business of finding and analyzing suitable episodes in a convenient way not only involves a lot of work, but is also bound up with subjective influences. Based on official match data, Fraunhofer IIS makes it possible to evaluate episodes objectively using performance parameters that have been defined by the customer. In addition, the characteristic attributes of these variables can be visualized in customized ways and the corresponding videos edited automatically. With our 10-plus years of experience in the processing of tracking data and our expertise in the fusion of tracking and video data, we are available to customers both for long-term collaborations in the form of a development partnership and also for product-oriented solutions. Our individual offers are aimed particularly at sports clubs, associations, league and media representatives and the manufacturers of league-wide tracking systems.



Löffler, C.; Mutschler, C.; Philippsen, M.: Approximative Event Processing on Sensor Data Streams (Best Poster and Demostration Award). In ACM (Hrsg.): Proceedings of the 9th International Conference on Distributed Event-Based Systems (9th ACM International Conference on Distributed Event-Based Systems, Oslo, Norway). 2015. S. 360—363. – ISBN 978-1-4503-3286-6.


Mutschler, C.; Philippsen, M.: Adaptive Speculative Processing of Out-of-Order Event Streams. In: ACM Transactions on Internet Technology (TOIT) 14 (2014), Nr. 1, S. 4:1—4:24, ISSN 1557-6051.


Mutschler, C.: Latency Minimization of Order-Preserving Distributed Event-Based Systems. Dissertation. Dr. Hut Verlag. 229 Seiten. ISBN 978-3-8439-1472-7.


Mutschler, C.; Löffler, C.; Witt, Nicolas; Edelhäußer, T.; Philippsen, M.: Predictive Load Management in Smart Grid Environments (Best Paper Award). In ACM (Hrsg.): Proceedings of the 8th ACM International Conference on Distributed Event-Based Systems (8th ACM International Conference on Distributed Event-Based Systems, Mumbai, India). 2014. S. 282—287. – ISBN 978-1-2734-4.


Mutschler, C.; Ziekow, H.; Jerzak, Z.: The DEBS 2013 Grand Challenge. In: ACM (Hrsg.): Proceedings of the 7th ACM International Conference on Distributed Event-Based Systems (7th ACM International Conference on Distributed Event-Based Systems, Arlington, Texas, USA)., 2013, S. 289-294. - ISBN 978-1-4503-1758-0.


Mutschler, C.; Witt, N.; Philippsen, M.: Do Event-Based Systems have a Passion for Sports?. In: ACM (Hrsg.): Proceedings of the 7th ACM International Conference on Distributed Event-Based Systems (7th ACM International Conference on Distributed Event-Based Systems, Arlington, Texas, USA)., 2013, S. 331-332. - ISBN 978-1-4503-1758-0.


Mutschler, C.; Philippsen, M.: Reliable Speculative Processing of Out-of-Order Event Streams in Generic Publish/Subscribe Middlewares. In: ACM (Hrsg.): Proceedings of the 7th ACM International Conference on Distributed Event-Based Systems (7th ACM International Conference on Distributed Event-Based Systems, Arlington, Texas, USA)., 2013, S. 147-158. - ISBN 978-1-4503-1758-0.


Loeffler, C.; Mutschler, C.; Philippsen, M.: Evolutionary Algorithms that use Runtime Migration of Detector Processes to Reduce Latency in Event-Based Systems. In: IEEE Computer Society (Hrsg.): Proceedings of the 2013 NASA/ESA Conference on Adaptive Hardware and Systems (AHS-2013) (2013 NASA/ESA Conference on Adaptive Hardware and Systems (AHS-2013), Torino, Italy)., 2013, S. 31-38. - ISBN 978-1-4673-6381-5.


Mutschler, C.; Philippsen, M.: Dynamic Low-Latency Distributed Event Processing of Sensor Data Streams. In: GI (Hrsg.): Proceedings of the 25th Workshop on Parallel Systems and Algorithms (PARS 2013) (25th Workshop on Parallel Systems and Algorithms (PARS 2013), Erlangen, Germany)., 2013, S. -.


Mutschler, C.; Philippsen, M.: Runtime Migration of Stateful Event Detectors with Low-Latency Ordering Constraints. In: IEEE (Hrsg.): Proceedings of the 2013 IEEE International Conference on Pervasive Computing and Communications Workshops (9th International Workshop on Sensor Networks and Systems for Pervasive Computing, San Diego, CA, USA)., 2013, S. 609-614. - ISBN 978-1-4673-5076-1.


Mutschler, C.; Philippsen, M.: Distributed Low-Latency Out-of-Order Event Processing for High Data Rate Sensor Streams. In: IEEE Computer Society (Hrsg.): Proceedings of 27th International Parallel and Distributed Processing Symposium (27th IEEE International Parallel & Distributed Processing Symposium (IPDPS), Boston, Massachusetts, USA)., 2013, S. 1133-1144. - ISBN 978-0-7695-4971-2.


Mutschler, C.; Philippsen, M.: Learning Event Detection Rules with Noise Hidden Markov Models. In: Benkrid, K.; Merodio, D. (Hrsg.): Proceedings of the 2012 NASA/ESA Conference on Adaptive Hardware and Systems (AHS-2012) (2012 NASA/ESA Conference on Adaptive Hardware and Systems (AHS-2012), Nuremberg, Germany)., 2012, S. 159-166. - ISBN 978-1-4673-1914-0.


Mutschler, C.; Philippsen, M.: Towards a Distributed Self-Optimizing Event Processing System for Realtime Locating Systems (RTLS). In: ACM (Hrsg.): DEBS PhD Workshops, 6th ACM International Conference on Distributed Event-Based Systems (International Conference on Distributed Event-Based Systems (DEBS), Berlin, Germany)., 2012.


Mutschler, C.; Kókai, G.; Edelhäußer, T.: Online Data Stream Mining on Interactive Trajectories in Soccer Games. In: Artesis University College of Antwerp (Hrsg.): Proceedings of the 2nd International Conference on Positioning and Context-Awareness (PoCA 2011, Brussels)., 2011, S. 15-22. - ISBN 978-94-90705-04-6.