The project began with open brainstorming around the concept of adaptive, evolving emergency response. Early sketches explored the idea of a shelter or system that could sense its environment, adapt according to triggers, and give an intelligent response. Ideas flew freely: a jacket that transforms, an aid kit, a learning kit, a shelter that changes form. The central question driving everything was: can it sense the environment, evolve according to the trigger, and give an adaptive response?
Shaping the Intelligence
From the chaos of early ideas, the team began structuring the underlying logic. They mapped out a simple but powerful input/output model:
Inputs: environmental changes, triggers (like fear/survival instinct), signals for harm
- Inputs: environmental changes, triggers (like fear/survival instinct), signals for harm
Outputs: an evolving emergency response — an urge to protect, act, or help
- Outputs: an evolving emergency response — an urge to protect, act, or help
Tools being considered included swiss army knife-style tent modules and modular survival shelters. The intelligence layer was taking shape.
The Final Idea — The Peace Missile / Aid CannonThe concept that won out was bold and provocative: an Aid Cannon — reframed as a Peace Missile. It senses climate catastrophes and fires relevant aid drops to those in need. The visual language was deliberate: a bundle of aid missiles launching parachuted supply packages. Something that looks like a weapon but delivers care. The name itself carries the statement.
Tasks, Plans & Tools — Moving AheadWith the concept locked, the team shifted to execution. A working list was drawn up across three pillars:
- N8N workflow Code + Arduino — gathering tangible data, troubleshooting, improving scraping and sourcing
- Physical Model — designing rockets/launchers, making physical prototypes for flood, earthquake, and fire scenarios
- Functionality — testing robots/launchers, writing code, putting it all together.
The Process
Creating the Automated system
The first step was gathering all relevant data that could be scraped and read to categorise different disasters and their scale. We explored several of the available weather widgets within n8n, but none contained the data we needed.
The next option was linking the system to multiple websites — pulling separate data streams for rainfall, seismic activity, droughts, and forest fires — then connecting that to an AI layer that reads the data and, based on set parameters, categorises the impact of each disaster. Depending on the impact level, it then triggers outputs to the respective Peace Missiles assigned to floods, earthquakes, or droughts.
However, feeding the AI data from multiple websites simultaneously proved to be too much to scrape and process effectively. So we simplified — and gave it one.
With the analysis side resolved, the next challenge was connecting everything to the Arduino — translating the data we were reading and analysing into a tangible physical result. This came with its own set of speed bumps, particularly getting n8n to communicate with the Arduino correctly. After some troubleshooting, we finally managed to bridge the connection through a shared WiFi network.
Designing the “missiles”
Multiple iterations of the missile were made, each attempting to shape the design around the specific aid it carried while deliberately distancing the aesthetic from military weaponry. However, complex geometries at such a small scale created countless issues with the 3D printer. The final design landed on something simpler and more aerodynamic. Given more time, we would have worked out the kinks and pushed the designs further. We also ran into some issues with the 3D printing, and so reverted back to a simpler design for the prototype.
Launching Mechanism
This was the stage where we significantly underestimated how long testing different mechanisms would take. We were fairly certain about the launch method early on — using the tension of a rubber band to propel the Peace Missile — but the real challenge was automating that launch via Arduino rather than triggering it by hand.
The key question became: how do you use a servo motor to release a rubber band at precisely the right moment, and what support structure does that require?
After several sketches and physical models, we arrived at a solution. A servo motor was positioned at a measured distance from the missile, connected to a rod-like key that passed through a wooden block with a corresponding hole. When the Arduino sent its input signal, the servo would rotate — pulling the key, which was attached to one end of the rubber band. This released the tension, causing the launch base to spring upward and send the 3D printed Peace Missile into the air.
The Project In ActionWe presented the project at a Work In Progress Exhibition at Elisava School of Design, Barcelona, where we shared our process and demonstrated the project. We received a lot of good feedback and also made some connections. Here is the project in action.
Cognitive Contribution LabelThe Cognitive Contribution label is an act of transparency showing how we used AI during the project development phases. (AI was a part of our n8n workflow which is separate from this)


_ztBMuBhMHo.jpg?auto=compress%2Cformat&w=48&h=48&fit=fill&bg=ffffff)





_vNic1Kho7Y.jpg)





Comments