Elite Sport Tracks the Squad.
Twinspire Studies the Individual.
You're Making Sunday's Decision with Tuesday's Data.
The Unresolved Decision
GPS in one platform. Strength data in another. Clinical notes remain local. Wearable signals often never leave the athlete’s own device. The challenge in elite sport is no longer access to data alone, but the absence of a coherent longitudinal model of the individual athlete across contexts and over time.
The Unresolved Decision
GPS in one platform. Strength data in another. Clinical notes remain local. Wearable signals often never leave the athlete’s own device. The challenge in elite sport is no longer access to data alone, but the absence of a coherent longitudinal model of the individual athlete across contexts and over time.
The Return-to-Play Gap
Football injury epidemiology is well established through long-running cohort studies, including the UEFA Elite Club research, which spans thousands of players across multiple countries and seasons. Yet despite increasingly sophisticated monitoring environments, the central decision problem remains unresolved: how to interpret changing signals at the individual level when decisions about return to training, return to play, and load modification must be made under uncertainty.
The Return-to-Play Gap
Football injury epidemiology is well established through long-running cohort studies, including the UEFA Elite Club research, which spans thousands of players across multiple countries and seasons. Yet despite increasingly sophisticated monitoring environments, the central decision problem remains unresolved: how to interpret changing signals at the individual level when decisions about return to training, return to play, and load modification must be made under uncertainty.
Recurrent Injuries
Subsequent and recurrent injuries continue to represent a meaningful part of the burden in professional football, and the period following return remains especially sensitive. The gap is therefore not simply one of data collection, but of individualized interpretation.
Recurrent Injuries
Subsequent and recurrent injuries continue to represent a meaningful part of the burden in professional football, and the period following return remains especially sensitive. The gap is therefore not simply one of data collection, but of individualized interpretation.
From Population Averages to Individualized Dynamics
Digital twin framework
Twinspire is developing a research-based digital twin framework for individualized athlete modeling. The system integrates longitudinal data from multiple sources, including training load, neuromuscular testing, clinician input, and wearable-derived physiological measures.
Approach
The approach combines principles from computational motor control, adaptive nonlinear systems, system identification, longitudinal sequence modelling, and self supervised learning to identify individualized representations of how an athlete’s physiological and functional state changes over time.
Structured data layer
In parallel, Twinspire is building a structured data layer that links sessions, tests, symptoms, and contextual information into a portable athlete history, supporting continuity and decision-making across performance and rehabilitation.
From Research Prototype to Applied Validation
A research prototype has been developed across mobile and web environments, designed to integrate heterogeneous data with minimal input from practitioners. The system now enters real-world validation, focusing on whether it remains usable, robust, and interpretable within elite performance settings.
The current work addresses three questions:
Are workflows operationally usable for staff?
Do data pipelines provide sufficient quality and continuity?
Do the signals support meaningful individualized modeling in practice?
Most systems describe load. Twinspire is being developed to model individual response.
Individualized Modeling
Most systems describe load. Twinspire is being developed to model individual response. The research draws on adaptive systems and individualized state estimation to understand how each athlete responds over time.
Multi-Signal Analysis
The goal is to establish a personalized reference state and detect meaningful deviations from it. The methodology investigates how multimodal data can be used to model evolving physiological and neuromuscular dynamics.
Validation & Risk Reduction
The current phase focuses on validating the approach and resolving key uncertainties before broader deployment.
Current Research Questions
- 01
How early can deviations from baseline be detected?
- 02
What data is required to establish a reliable individualized model?
- 03
How do missing data and changing contexts affect robustness?
- 04
Can compensation patterns be identified before symptoms emerge?
Who We Work With
We are currently reaching out to clubs, clinics, and performance institutions to participate in our research validation study.
Footballers
Your body has been generating data your whole career. Most of it has disappeared. This research aims to establish whether a structured data passport can change that and give you ownership of your own athletic history.
Coaches & Performance Staff
Help us understand how readiness modeling integrates into real decision-making environments, and where the model's outputs are and aren't actionable.
Physiotherapists
Co-develop unified deviation-alert and recovery-signal workflows grounded in multi-source data. Help us understand what a clinically useful signal actually looks like in practice.
The Research Came First. The Team Was Built Around It.
The project is supported by a dedicated research team of four based at DTU, contributing specialisms in physiological modeling, data engineering, biomechanics, and neuromuscular analysis. Their work forms the scientific backbone of the validation studies currently underway.
Athlete Data Must Be Handled and with Responsibly, Transparently, Clear Boundaries.
Twinspire is being developed with data governance and research ethics as core design principles.
Core Principles
The current phase is focused on scientific validation, with attention to lawful processing, traceability, transparency, and responsible handling of longitudinal athlete data.
Governance & Compliance
The project is built around GDPR-aligned data practices, EU-based hosting and processing, relevant data processing agreements, and research workflows designed to support clear governance across clubs, practitioners, and athletes. Where appropriate, research data management is structured in accordance with FAIR principles.
Research Disclaimer
Twinspire is a research-stage decision support prototype. It is not intended to diagnose, treat, or replace clinical judgment.
FAQ
Answers to the most common questions about Twinspire, our methodology, and how clubs can get started.
Twinspire is an AI-powered decision support platform for football clubs. It builds a continuously updated physiological model of each player by drawing on GPS data, strength output, neuromuscular sensor data, physio notes, and the athlete's own wearables, and alerts medical and coaching staff when something shifts. The goal is to catch the warning signs before an injury becomes a reinjury.