ACHIEVEMENTS TO DATE

In just two years, BIGDATPOL has taken important steps to establish itself as a European reference framework for responsible big data policing. The first milestone has been the creation of a comprehensive database of initiatives across Europe, systematically mapping which types of data are used, how they are analysed, and in which organisational contexts they take shape. This resource reduces fragmentation and provides a basis for comparison, evaluation, and knowledge sharing across countries.

Building on this effort, BIGDATPOL also produced the first European typology of big data policing methods. This typology provides a conceptual framework for assessing opportunities and risks across diverse approaches, and it already serves as a reference in both academic and policy debates.

Achievements have not been limited to infrastructure and data. BIGDATPOL has also succeeded in mobilising a broad network of stakeholders. The project now connects hundreds of academics, practitioners, and policymakers through workshops, interviews, and online platforms. This network ensures that the programme’s work is closely aligned with practice and that insights are shared in both directions.

Finally, BIGDATPOL has gained visibility well beyond academia. With invitations to present at Europol, recognition during Belgium’s EU Council Presidency, and selection for the Docville Science Pitch festival, the project has already contributed to public debate on the ethical and societal dimensions of artificial intelligence in policing.

These achievements demonstrate the breadth of progress made in a short time: building infrastructures, providing conceptual clarity, creating empirical foundations, strengthening networks, and raising societal awareness. Each of these strands feeds into the same ambition of showing how data-driven policing can be developed responsibly, transparently, and with public legitimacy.

BEYOND THE STATE OF THE ART

BIGDATPOL does not simply add another set of studies to the growing field of predictive policing. It deliberately moves the field beyond small-scale pilots, ad hoc experiments, and retrospective analyses that have often characterised the debate. By combining large-scale empirical testing with a strong interdisciplinary and ethical foundation, the programme is setting new standards for how innovation in policing should be designed, evaluated, and governed.

Preparations are underway for the largest coordinated experimental trials of big data policing in Europe. These trials will not only examine whether predictive tools can improve the allocation of resources, but will also assess wider outcomes such as legitimacy, trust, proportionality, and cost-effectiveness. This ensures that effectiveness is never separated from questions of accountability and societal value.

A second distinctive feature of BIGDATPOL is its integrated framework. Criminological theory, statistical modelling, legal expertise, ethical reflection, and economic assessment are not treated as separate tracks but as interconnected dimensions of one approach. This makes it possible to evaluate predictive policing initiatives not only on their accuracy, but also on their transparency, compliance with fundamental rights, and real-world impact.

Finally, BIGDATPOL stands out by its scalability. The same framework is applied across different domains, from burglary to fraud, football policing, drug trafficking, cybercrime, and environmental offences, showing that responsible data-driven approaches can be generalised without losing sensitivity to local contexts. By connecting fragmented initiatives under one coherent methodology, the programme paves the way for interoperability and long-term learning at the European level.

METHODOLOGICAL INNOVATIONS

A central ambition of BIGDATPOL is to ensure that advances in data science go hand in hand with criminological theory and ethical responsibility. To achieve this, the programme has introduced several methodological innovations that significantly extend the analytical toolkit available for studying crime and security.

One of these is the “crime heartbeat” methodology, which captures temporal and spatiotemporal concentrations of crime in ways that reveal recurring rhythms of criminal activity. This approach allows for more precise modelling of when and where risks are likely to intensify.

Another innovation is the integration of crime scripting into predictive models. By incorporating insights from situational crime prevention, BIGDATPOL connects the “how” of criminal opportunity structures to the “where” and “when” of predictive analytics, resulting in models that are both more interpretable and more practically useful.

The programme also pioneers a hybrid approach to data, combining conventional sources such as census and administrative statistics with novel ones including mobile phone data, Google Places, and even Google Street View imagery. This not only enriches predictive validity but also raises new ethical and legal questions, which are explicitly addressed in the research design.

Finally, BIGDATPOL has conducted systematic comparisons of machine learning algorithms across European contexts. By benchmarking their performance, transparency, and susceptibility to bias, the project provides robust evidence on which techniques are best suited for responsible applications in policing.

These methodological innovations demonstrate how advanced analytics can be reconfigured to serve not just predictive accuracy, but also transparency, accountability, and theoretical grounding. They are among the key contributions that make BIGDATPOL an international reference point in the debate on artificial intelligence and security.