A real-time disaster response management game inspired by American Red Cross tabletop exercises. Open shelters and kitchens, manage staff, dispatch vehicles, and make high-stakes decisions to support displaced communities — under tight budgets and time pressure.
Gameplay Preview
Check out our demo video for a look at the gameplay.
What Is C.O.R.A
C.O.R.A is a real-time disaster response management game inspired by American Red Cross tabletop exercises. Players open shelters and kitchens, manage staff, dispatch vehicles, and make high-stakes decisions to support displaced communities, under tight budgets and time pressure.
Open shelters and kitchens, manage staff, and dispatch vehicles to support displaced communities — all under tight budgets and time pressure.
Core GameplayAI teammates with distinct personalities propose actions, surface problems, and challenge player choices, making human-AI collaboration central to gameplay.
AI LayerOrganizers can customize maps and disaster conditions, ensuring every scenario feels dynamic, replayable, and unpredictable.
Admin SystemPlayers learn how staffing, logistics, shelter capacity, budgets, and community needs interact under pressure — failures in one area quickly affect the rest.
Learning DesignAI teammates introduce negotiation and delegation challenges, requiring players to assess recommendations, manage differing priorities, and decide when to trust automation.
Human-AI TeamingThe AI agent layer is partially integrated, with a mix of LLM-proposed and template-based tasks. Full integration is in progress.
In DevelopmentLearning Goals
C.O.R.A teaches systems thinking through cascading consequences rather than direct instruction. Players learn how staffing, logistics, shelter capacity, budgets, and community needs interact under pressure — failures in one area quickly affect the rest.
Configurable scenarios allow facilitators to tailor learning goals across skill levels and disaster contexts.
Real-time resource allocation promotes planning, prioritization, and decision-making under uncertainty, forcing players to balance efficiency and human impact.
AI teammates introduce negotiation and delegation challenges, requiring players to assess recommendations, manage differing priorities, and decide when to trust automation.
Configurable scenarios allow facilitators to tailor learning goals across skill levels and disaster contexts.
Accessibility
C.O.R.A is designed for diverse learners and facilitators. We're committed to lowering barriers so that more people can participate in disaster preparedness education.
Right-clicking on any in-game text triggers live speech audio, supporting players who benefit from auditory feedback during gameplay.
Runs directly in any modern browser via WebGL — no downloads, no setup, no IT barriers. Accessible from any device with a network connection.
A dedicated admin interface lets organizers build and save custom scenarios without coding knowledge, ensuring instructors can tailor the experience to their learning objectives.
Scenario text, UI labels, and agent dialogue are stored in configurable data files, making localization straightforward for facilitators in non-English contexts.
Who We Are
C.O.R.A. is developed at Carnegie Mellon University, through the NSF AI Institute for Societal Decision Making (AI-SDM), in collaboration with the American Red Cross.
Ready to Respond?
Play a disaster response management game where every decision matters. Try it in your browser!
Note: C.O.R.A is a work in progress. The public build is available for general use. The instructor configuration system is intended for facilitators only, and full LLM agent integration is still under active development.
Try Our Game