Research Centre
AI-Driven Early Warning for Community Violence
The Core Philosophy: From Reactive to Proactive
Current violence prevention systems often suffer from "Reactive Lagging"—we respond only after an incident has occurred. Our research, presented at the 31st German Prevention Congress (DPT) in Hannover, proposes a fundamental paradigm shift toward "Proactive Leading".
Unlike fictional "pre-crime" models that target individuals, our system targets friction. We are mapping places and networks, not individual faces, to identify where a community may need additional support before a conflict escalates.
Our Research Mission
At the EBS Centre, our mission is to transform community safety by transitioning from "Reactive Lagging" to "Proactive Leading." We believe that the future of public safety lies not in increased surveillance of individuals, but in the intelligent understanding of environments and social dynamics.
Bridging Technology and Practice
Our research is dedicated to developing AI-augmented early warning mechanisms that empower practitioners—rather than replace them. By integrating high-fidelity, anonymized data with social sciences, we aim to bridge the "research-practice gap" in community violence prevention.
The Three Pillars of Our Mission:
-
Targeting Friction, Not People: Our primary goal is to map places and networks, not individual faces. We apply Situational Action Theory (SAT) and social contagion models to identify volatile "Settings" where community support is needed most.
-
Ethical Innovation by Design: We are committed to a "Non-Punitive" framework. Our mission is to ensure that AI alerts trigger social outreach and community engagement, strictly avoiding "pre-crime" enforcement traps.
-
Radical Transparency: Through the use of Explainable AI (XAI) and our Integrated Data Platform (IDP), we ensure that every algorithmic insight is understandable and verifiable by human experts.
Theoretical Architecture: The "Brain"
Our prototype merges two leading sociological perspectives to create a robust Setting-Based Risk Score, which acts as a legal safeguard against individual profiling:
-
Social Contagion (Papachristos): We view violence as a network-based phenomenon, identifying the ties through which tension flows between groups.
-
Situational Action Theory (SAT) (Wikström): We identify the "Setting"—the specific environmental triggers and social contexts—that make a particular time or place volatile.
The Technical Pipeline & Operational Workflow
The system is built on a sophisticated Integrated Data Platform (IDP) that ensures privacy through a three-step process: Ingestion, Standardization, and Anonymization.
Data Streams & Analysis
The platform synthesizes historical incident reports, community social indicators, and real-time (non-PII) mobility data. Using Pattern Detection, the AI identifies correlations often missed by human analysis to assess the volatility of specific "Nodes".
Human-in-the-Loop (XAI)
Transparency is central to our design. Our Early Warning Dashboard utilizes Explainable AI (XAI). Rather than a vague "danger" alert, the system explains the underlying logic (e.g., "30% increase in transit density combined with historical friction").
The Practitioner Feedback Loop ensures that human expertise always overrides the algorithm. Social workers and outreach teams validate alerts; if a "False Positive" occurs, that data is used to retrain and refine the model.
Ethical & Legal Defense: Our "Shield"
-
GDPR Compliance: All data is handled via aggregation and strictly non-PII mobility layers.
-
Non-Punitive Design: Alerts trigger prevention action (outreach and community engagement) rather than police enforcement units.
-
Decision Support: The system is designed as a tool for practitioners, ensuring human agency remains the final authority in any intervention.
Call for Participation: Join Our Research
We are actively seeking to bridge the gap between technological innovation and practical prevention. Preliminary pilot results have already demonstrated how AI outputs can be successfully integrated into local prevention networks.
We invite interested researchers, practitioners, and technologists to collaborate on the next phase of this project. If you are interested in implementation challenges, ethical governance, or cross-sector collaboration in AI-driven prevention:
Apply to participate at: Contact us
Humanizing the Algorithm: A Global Research Initiative on AI-Driven Work & Well-Being
The Challenge
As Artificial Intelligence (AI) becomes deeply embedded in hybrid and asynchronous work, the corporate world faces a critical contention: will these tools foster "human flourishing" or become new instruments of digital control?. Current transitions are often uneven, leaving a significant gap in management literature regarding how to intentionally structure AI-enhanced models to improve organizational happiness while remaining ethically and legally aligned.
Our Research Mission
Led by the Education Beyond Science (EBS) Centre, this project moves "beyond science" to synthesize technology, law, business and psychology. Using the Job Demands–Resources (JD-R) model and Self-Determination Theory (SDT), we are empirically testing how AI can be transformed from a source of "algorithmic anxiety" into a supportive resource for employee agency.
The Multi-Pillar Framework
We invite contributors to engage with our four-pillar model for sustainable digital management:
-
AI-Enhanced Flexibility: Evaluating intelligent scheduling and task automation.
-
Legal & Ethical Safeguards: Examining the "Right to Disconnect" and algorithmic transparency as buffers against burnout.
-
Organisational Dynamics: Shifting management from surveillance to trust-based psychological safety.
-
Human Outcomes: Measuring the ultimate success of AI through stress reduction and organizational happiness.
Our Methodology & Data
This study leverages the world's most comprehensive longitudinal datasets to ensure global validity:
-
EWCS: Tracking job autonomy and digital tool usage.
-
OECD: Mapping country-level governance and employment quality.
-
Gallup World Poll: Measuring the engagement and thriving of the global workforce.
Call for Participation: Humanizing the Algorithm
We are actively seeking researchers, policymakers, and organizational leaders to collaborate on this global empirical validation of AI-driven work models. Our initiative investigates how the integration of Artificial Intelligence impacts the fundamental pillars of human well-being and professional agency.
Whether through data sharing, policy analysis, or pilot implementations, your expertise will help define a future where technological advancement and societal progress move in tandem.
How to Get Involved
-
Empirical Validation: Contribute to our ongoing studies on the psychological and social impacts of algorithmic management in the workplace.
-
Policy & Ethics Advocacy: Help us translate research findings into actionable policy frameworks that protect worker autonomy and mental health.
-
Organizational Pilots: Partner with us to implement and test "Human-in-the-Loop" work models that prioritize employee well-being alongside efficiency.
-
Data Collaboration: Share anonymized organizational data to help us identify the specific "friction points" where AI integration affects team cohesion and individual stress.
Explore the Research Foundation
-
Review the Statistical Analysis Plan (SAP): Dive into our methodology, including Multilevel Modelling (HLM)and Structural Equation Modelling (SEM) pathways used to map the complex relationships between AI tools and human outcomes.
-
A Shared Vision: Partner with the EBS Centre to ensure that the next generation of work models remains centered on human dignity and professional growth.
Join our global research initiative: Contact us