NERV | Devpost

Inspiration: Cities are full of problems that everyone can see, but nobody connects. A flooded footpath, a broken streetlight, an overflowing bin, a dangerous crossing, or an inaccessible road may each seem like isolated incidents. Usually, a city only reacts when enough people complain or when the problem has already become a crisis. That made us ask a simple question: What if a city could feel its own problems before they became emergencies? The human body does not wait for one neuron to send a formal complaint before reacting to pain. Millions of tiny signals travel through a nervous system, allowing the body to understand what is happening and respond. NERV brings that idea to cities. Instead of treating every citizen report as an isolated complaint, NERV turns everyday observations into signals that can collectively reveal emerging urban problems.

What it does: A citizen should not need to understand which government department is responsible for a problem or fill out a complicated form. With NERV, they can simply share what they see through a photo, voice note, or short description. Gemini understands the unstructured information: what the problem is, where it is happening, who may be affected, and how serious it appears to be. But the real idea goes beyond reporting. NERV connects many small and seemingly unrelated signals across time and location. For example, a few photos of standing water, voice notes about mosquitoes, reports of damaged drainage, and pedestrians avoiding the same route may individually look like unrelated complaints. Together, they may reveal something much more important: An emerging drainage and public-health risk that should be addressed before it escalates. NERV transforms isolated observations into an Emerging Issue and helps identify the smallest meaningful intervention before the city is forced into expensive, reactive action.

How Gemini brings NERV to life: Gemini is the intelligence that allows NERV to understand the messy way humans naturally communicate. It can help interpret: Images of urban problems Voice notes in natural language Text descriptions Context about affected people and locations Gemini can then transform these observations into structured signals, identify common patterns, reason about possible connections, and help visualize how a small problem could develop into a larger one. For example: Individual signals Standing water → Damaged drain → Pedestrians avoiding crossing → Mosquito complaints

NERV's interpretation: Emerging urban risk: drainage failure with potential mobility and public-health impact.

What we learned: While developing NERV, we realized that cities do not necessarily lack information. They lack the ability to connect weak signals early enough. People already know when their neighbourhood is beginning to fail. They see the small changes long before those changes become official statistics, emergency calls, or headlines. The challenge was imagining how AI could understand these observations without forcing citizens to become data-entry operators. That led us to Gemini's multimodal capabilities. Instead of designing another complicated civic reporting system, we imagined an interface where people communicate naturally and AI does the difficult work of understanding and connecting the signals.

Challenges: The biggest challenge was balancing an ambitious vision with a realistic path to implementation. A system that claims to predict every urban problem would be unrealistic. NERV instead starts with a practical approach: identify patterns that humans might miss when information is fragmented across thousands of individual observations. Privacy was another important consideration. NERV should focus on understanding urban conditions rather than building profiles of individual citizens. The system can aggregate and anonymize signals while still preserving the geographic and contextual information necessary to identify emerging issues.

The vision: Today, cities are largely reactive. A problem happens. People complain. Authorities investigate. Action comes later. NERV imagines a different future. A city that can listen continuously to the people living inside it. A city that can recognize weak signals before they become visible crises. A city that can feel.