Fourth-generation deterioration modelling

Model how infrastructure ages —
and what it costs to keep going.

Juno Cassandra is a domain-agnostic framework for simulating the deterioration of roads, bridges, water networks and powerlines. Bring your own logic, optimise treatments with MCDA or goal-seeking, and let the data help drive your decisions.

Read the docs

What it does

A framework you extend — the modelling plumbing is done, the domain logic is yours.

Domain agnostic

Roads, bridges, water, powerlines — one engine. Plug in your own domain model and the framework does the looping.

R-powered reporting

Upload your own R scripts, or pick from our pre-built kits — they run server-side against your outputs and return report-ready charts and tables you can drop straight into a deliverable.

Optimisation built in

Multi-Criteria Decision Analysis and Incremental BCR for treatment selection. Goal-seeking inverts the question — what budget hits the KPI?

ML-ready

Drop ML.NET models alongside equation-based rules — random forests, SVMs, your own predictors, scored at runtime.

Debuggable by design

Informative error messages, warning detection, offline debugging. Things go wrong; the framework helps you find out where.

Desktop or web

Run the WinForms toolbox on your machine, or the hosted web app at junocassandra.com. One framework underneath.

One framework. Many networks.

  • Roads
  • Bridges
  • Water networks
  • Powerlines

From raw asset data to defensible decisions.

Cassandra sits between your asset register and the decisions you need to justify — maintenance plans, renewal budgets, intervention thresholds.

  • Five-stage modelling cycle: initialise, trigger, maintain, reset, deteriorate — over any horizon you specify.
  • Compare budget scenarios; track KPIs across modelling periods.
  • Run R scripts on outputs server-side — your own, or pre-built kits that ship report-ready charts and tables.
  • Monte-Carlo simulation for risk analysis (desktop toolbox today; web is roadmap).
  • Outputs as CSV — your data, your formats, no lock-in.
run.log 
12:04:18 info  loading workbench → ok
12:04:19 info  pavement domain model v3.1
12:04:22 info  optimisation: incremental-bcr
12:06:47 warn  elements > 90 % at yr 18 → 3.2 %
12:07:03 info  writing outputs/run_baseline/
12:07:05 done  complete — 2m 47s