Inspiration

Cities operate dozens of interdependent infrastructure systems — power grids, water treatment, transit networks, waste processing, emergency services — but the people responsible for keeping them running rely on siloed dashboards and spreadsheets. When a water main breaks, it affects power cooling, which affects transit signals, which affects emergency response times. These cascade relationships are invisible in traditional tooling.

We wanted to build the interface that municipal operators should have: a
single graph-based view where every system, dependency, and failure chain is visible at a glance.

## What it does

AXIOM models a city's infrastructure as an interconnected node graph. Each
node represents a system or subsystem (power substation, water treatment plant, transit hub). Edges represent dependencies, resource flows, and data
connections between them.

The platform provides:

Real-time health scoring — efficiency metrics per node and aggregate
scoring across the full graph
Anomaly detection — automated identification of degraded systems and
emerging failure patterns
Cascade risk analysis — explicit modeling of how a failure in one system propagates through dependencies
Optimization recommendations — scored improvement suggestions ranked by cost-effectiveness and projected impact

Everything renders in a blueprint-grid interface designed for high information density and rapid situational awareness.

## How we built it

Frontend: Next.js 14, React, Tailwind CSS
Visualization: Custom node graph renderer with GPU-accelerated layout (Cosmograph/D3.js pipeline)
Design system: Blueprint-grid dark theme — Share Tech Mono and Rajdhani typography, cyan-dominant palette (#4ab8d8), technical grid overlay inspired by engineering schematics
Scoring engine: Continuous evaluation loop computing node health, edge
stress, and aggregate system efficiency
Data model: Nodes carry metadata (capacity, age, condition scores); edges carry dependency type and criticality weight

We started with a single-file HTML prototype to validate the full UI concept — interactive graph, stat panels, detail views — before decomposing into the production stack.

## Challenges we faced

Graph layout at scale — Positioning dozens of interconnected nodes without visual clutter required iterating on force-directed layouts and manual
clustering by infrastructure domain. Readability at a glance was
non-negotiable.

Cascade modeling — Real infrastructure dependencies aren't simple
parent-child trees. They're cyclic, weighted, and context-dependent. A power failure affects water pumping, but water cooling also affects power
generation. Modeling bidirectional cascade risk without infinite loops
required careful graph traversal with decay functions.

Information density vs. clarity — Municipal operators need to see
everything at once, but a wall of data helps no one. The blueprint aesthetic wasn't just cosmetic — the grid, muted palette, and typography hierarchy are load-bearing design decisions that keep 12+ metrics readable simultaneously.

## What we learned

Node-edge graph models surface relationships that tabular data hides
completely
Blueprint/schematic visual language reduces cognitive load for technical
operators compared to consumer UI patterns
Prototyping the full UI in a single HTML file before touching a framework
saved significant iteration time
Municipal infrastructure optimization is a domain where software is decades behind the complexity of the systems it manages

What's next for AXIOM

Live data ingestion — Connect to real municipal APIs (SCADA, GTFS
transit feeds, utility MQTT streams) to replace static demo data with live
system telemetry
Cascade simulation — "What-if" mode where operators can kill a node and watch failures propagate through the graph in real time, quantifying blast
radius before incidents happen
Historical replay — Time-scrubber to review how infrastructure health
evolved over days/weeks, correlating anomalies with weather, events, or
maintenance windows
Multi-city benchmarking — Compare efficiency scores across
municipalities to surface best practices and infrastructure patterns that
scale
Alert routing — Push degradation warnings to the right department before cascades reach critical systems, with severity scoring based on graph
centrality