SafeFuture Lab

Ask the work, not the model

Every agent here is restricted to one published piece of road-safety analysis and its cited sources. Answers are machine-verified before you see them; when an answer can't be verified, the agent refuses. No general chatbot. No improvising.

Project-restricted · machine-verified · refuse-when-unverifiable Humans audit every interaction No agency endorsement implied — affiliations & provenance stated per work
The agents

Six published works, six verified assistants

RT Every number SQL-verified

Australia's Road Toll

A queryable report on the national road toll and Victoria's turn — rolling totals, state contributions, five-year trends, volatility, remoteness, and pre-registered falsification gates. Built on BITRE microdata.

roadtoll.crashstats.com.au → VF Every number SQL-verified

Vehicle Fatality Trends

Victoria’s fatal crashes by vehicle brand and model — raw involvements, fleet-adjusted rates, the vulnerability asymmetry (who dies when a model crashes), vehicle mass and age — with an assistant answering strictly from the report’s tables. Involvement, never fault.

vehicles.crashstats.com.au → SE Every number SQL-verified

Speed Enforcement Effect

NSW fixed cameras × connected-vehicle operating speeds on matched-grade pairs — the hill effect, the at-camera gap across 17 strict pairs, and the selection trap published as a warning — with an assistant answering only what the marts can verify. Descriptive, never causal.

enforcement.crashstats.com.au → RR Every answer fact-checked

Road Risk Explainer

The five ways to measure road-network risk — collective risk, personal risk, AusRAP Star Ratings, Infrastructure Risk Rating and the systemic method — with an assistant answering strictly from the published explainer and its primary sources.

ausrap.crashstats.com.au → SS Every answer fact-checked

Safe System Assessment

The Safe System Assessment framework explained — matrices, scoring, and how SSA sits alongside other road-safety evaluation tools — with a project-restricted assistant on the published knowledge base.

ssa.crashstats.com.au → NM Every answer fact-checked

Near-Miss × Crash Divergence

Melbourne-metro surrogate-safety R&D note — a descriptive comparison of telematics near-miss density and crash patterns under explicit exposure controls: the aggregation paradox, candidate cells for further work, and an exploratory modelling-lift test. Assistant restricted to the published note. Candidates, not verdicts.

nearmiss.crashstats.com.au →
How these agents work

The rules every agent on this page obeys

  1. Project-restricted. Each agent answers only from one published work and its cited public sources — never from general model knowledge.
  2. Machine-verified. Numbers come from read-only queries against the same data that produced the published charts; a deterministic checker validates every figure in every answer before it is shown.
  3. Refuse rather than guess. Out-of-scope or unverifiable questions get an honest refusal, not an improvisation.
  4. Audited. Every interaction is logged and reviewed by a human; flagged answers are corrected or withdrawn.
  5. Disclosed bias, open data. Authors declare their interests; claims ship with falsification tests where applicable; sources are public and named.