AuNav — Deep-Dive Research & Development Plan (v2)¶
Product name: AuNav Prepared: 30 July 2026 · Revision: v2 (detailed, research-backed) Premise: Build a community-driven navigation app that matches Waze's core experience but is purpose-built for Australian conditions — especially truckies, long-haul/interstate drivers, and remote/outback travel. Consumers get a smarter, Australia-built nav; fleets get a compliance dashboard they'll pay for.
What changed in v2 (the four findings that reshape the plan): 1. A pure "Google Maps under the hood" build is not viable as originally drawn. Google's ToS prohibits rolling your own turn-by-turn by combining Directions + Maps SDK; compliant in-app navigation requires the Navigation SDK (Enterprise, negotiated). GMP also offers no developer offline mode and no native heavy-vehicle routing. → The architecture becomes dual-engine: Google online in metro, an OSM/HERE offline+truck engine for outback and heavy-vehicle routing. (Part 6, Part 11) 2. The authoritative AU heavy-vehicle network (NHVR National Network Map) is view-only + a protected API, not an open bulk download. Assemble a working dataset from state CC-BY open data + NHVR Spatial API + crowdsourcing for local-road low bridges. (Part 5) 3. A major HVNL reform lands 1 August 2026: BFM/AFM accreditation modules are replaced by Alternative Compliance Hours (ACH), with a new penalty schedule and updated work diary. Build the fatigue engine on work/rest limits (stable), not module names. (Part 8) 4. The market gap is specific and real: no product fuses AU truck-legal routing + outback/offline coverage + HVNL/EWD compliance + live fuel, and the ~64,700 small road-freight businesses / owner-drivers are underserved by hardware-heavy, quote-based incumbents (~$35+/veh/mo). That is the wedge. (Part 3, Part 12)
Part 0 — Executive summary¶
Australia has ~760,000 powered trucks, ~64,700 road-freight businesses (mostly owner-drivers and micro-fleets), a ~A$74bn road-freight industry, and >901,000 registered caravans/campervans feeding a ~A$18bn RV market. Every one of these drivers uses navigation that was built for cars in cities.
Waze is the closest analogue to what we want to feel like — crowdsourced alerts, live re-routing, speed-camera warnings — but it has no truck routing, no offline mode, thin rural coverage, and no fatigue/compliance awareness. Truck-specific apps (Sygic, CoPilot) route trucks but have weak AU outback and no HVNL compliance; outback apps (Hema) nail offline tracks but don't do truck legality or compliance; telematics incumbents (Teletrac Navman, EROAD, MTData) do compliance but are hardware-heavy, enterprise-priced, and driver-hostile. Nobody owns the intersection.
Our product sits at that intersection: a best-in-class consumer nav (free, ad-supported) with an AU brain — truck-safe routing, HVNL fatigue planning, offline outback resilience, one unified official-hazard layer, live fuel — and a B2B fleet-compliance SaaS (the revenue engine) priced to win the small operators the incumbents ignore.
Revenue model: two-sided. B2C freemium (~A$9.99/mo Pro) + Google-mediated ads; B2B fleet SaaS at A$15–40/vehicle/month undercutting incumbents, with a free-EWD acquisition hook.
The hard parts, named honestly: (1) dual-engine map/routing architecture and its cost; (2) assembling and maintaining the heavy-vehicle restriction dataset (the moat and the biggest effort); (3) safety/legal liability of "clear/safe" truck guidance and fatigue advice; (4) cold-start crowdsourcing in low-density areas. Each has a mitigation in this plan.
Part 1 — Deep dive: how Waze works & its full feature set¶
1.1 What Waze fundamentally is¶
Waze is a crowdsourced, community-first navigation app. Unlike Google Maps (authoritative data + traffic modelling), Waze's moat is real-time reports from millions of drivers feeding a live routing engine, plus a volunteer community that edits the map itself (Waze Map Editor, WME). Google owns Waze but keeps the products distinct — though they are steadily merging (Waze incident reports began appearing inside Google Maps in Nov 2024).
1.2 Waze feature inventory (what we must match)¶
A. Navigation & routing — turn-by-turn voice nav with continuous live re-routing; dynamic ETA; lane guidance; alternative routes with time/distance/toll trade-offs; multi-stop and planned drives (leave-by time); speed-limit display + over-speed warning; toll estimates and toll/highway/dirt avoidance; roundabout exit callouts (important in AU).
B. Crowdsourced real-time reports (the heart of Waze) — reportable events: traffic jam, crash, hazard (road/shoulder/weather), police, speed/red-light camera, road closure, blocked lane, stopped vehicle, object on road, pothole, construction, fuel prices. "Reports ahead on your route" proactive alerts; thumbs-up / "still there" / "not there" confirmation to age-out stale reports.
C. Community map editing — WME lets volunteers fix roads, turn restrictions, house numbers, closures; rank/permission tiers and local "area managers". This keeps maps fresh cheaply — a genuine differentiator vs licensing a static map.
D. Social & gamification — points, ranks, moods/avatars, leaderboards; send-ETA / live-location share; groups.
E. Personalisation & lifestyle — saved places, planned drives, calendar integration; cheapest fuel along route; parking suggestions; music/podcast controls; CarPlay/Android Auto; vehicle type affecting routing.
F. Safety — speed-camera & average-speed-zone alerts; sharp-curve, speed-bump, railroad-crossing warnings; "vehicles stopped on shoulder" alerts; school-zone & pedestrian-crossing alerts (added late 2024, time-of-day aware).
1.3 Waze in 2024–2026 (current, researched)¶
- Gemini-powered conversational reporting (testing from Oct 2024; extended Jul 2026): report hazards and map corrections by speaking naturally; spoken map-edit suggestions are routed to local map editors for verification before applying.
- Gemini destination search: natural-language POI search ("a gas station nearby with the lowest prices").
- Motorcycle Mode (Jul 2026): two-wheeler routing + hazard warnings (rolling out in LatAm/SE Asia first).
- Personalised/learned routes from your own driving history; "less chatty" mode for fewer prompts.
- Alert clustering (merging co-located alerts into one) — roadmapped/announced since Jan 2024 but not confirmed fully shipped as of mid-2025; treat as aspirational.
- EV routing (since Mar 2023): enter EV model + connector type → routes only to compatible chargers; charger data kept current by map editors.
- Google Assistant is being removed, not added — Waze discontinued Assistant on iOS in 2024 and is replacing it with its own Gemini voice stack. (Implication: voice UX is now a first-class, AI-driven surface — plan ours the same way.)
1.4 Waze's business model (researched — important nuance)¶
- Waze shut down its own self-serve ad marketplace on 31 Aug 2023. The classic "Waze Ads / Waze Local" self-serve dashboard no longer operates; advertising folded into Google Ads, primarily Performance Max for Store Goals, with Waze as one inventory surface ("Promoted Places in Navigation" pins). Most online "Waze Ads" guides describe the retired 2018–2023 model — treat those CPMs (Branded Pins ~$2 CPM, Zero-Speed Takeover ~$30 CPM, entry from ~$2/day) as historical/indicative.
- Waze for Cities (formerly Connected Citizens Program): a free, two-way data exchange with ~1,000 government/agency partners — Waze sends anonymised incident/traffic data; cities send official closures/roadworks/events. This is a strong reference model for a public-sector data-sharing arm of our product (and a channel to official AU hazard feeds).
1.5 Where Waze is weak in Australia today (our opening)¶
- No heavy-vehicle routing — ignores low bridges, weight/length limits, B-double/road-train networks.
- No real offline mode — Waze needs data; huge swathes of AU have none.
- Thin community coverage outside metros — crowdsourcing collapses exactly where help matters most.
- Doesn't integrate official AU hazard feeds (flood/fire/outback closures).
- No fatigue/rest-law awareness; fuel/roadhouse data sparse on remote highways.
Part 2 — Australia-specific problems worth solving¶
2.1 Heavy-vehicle / truck routing (highest-value, hardest)¶
- Low bridges & clearance strikes. NSW has more sub-4.6 m clearances than any state; QLD bridge strikes caused ~200 hours of road-user delay in a single recent year; individual strikes have caused $1M+ damage. Car GPS routes trucks straight into these. NSW penalty for ignoring low-clearance signage: $5,500 + 12 demerit points (up to $6,676 + 4 demerits if a structure is damaged).
- Restricted Access Vehicles (RAVs). B-doubles, B-triples and road trains may only travel on approved/gazetted networks (under notice) or under an NHVR permit (off-network). Off-network travel is an offence. Drivers currently juggle the NHVR Route Planner + permits + a separate nav app.
- Mass/dimension limits: length, height, width, axle mass, Higher Mass Limits (HML) routes, PBS levels, tunnel hazmat bans.
- Permit classes: Class 1 (oversize/overmass), Class 2 (B-doubles/road trains/PBS/HML), Class 3 (miscellaneous). Access = notice (self-serve on approved network) vs permit (apply via NHVR Portal; NHVR consults road managers → often slow).
2.2 Fatigue management & HVNL compliance (professional drivers)¶
- HVNL fatigue law applies to fatigue-regulated heavy vehicles (>12 t GVM, or combination >12 t, or bus >4.5 t carrying >12 adults) across ACT/NSW/QLD/SA/TAS/VIC. WA and NT run their own regimes but a WA/NT vehicle becomes HVNL-bound the moment it crosses into an HVNL state.
- Rulesets: Standard hours (max 12 h work / 24 h, min 7 h continuous rest), BFM (up to 14 h with stronger night rest + accreditation), AFM (risk-based, negotiated within outer limits). From 1 Aug 2026, BFM/AFM modules are replaced by Alternative Compliance Hours (ACH) under the new Heavy Vehicle Accreditation (HVA) scheme.
- Penalties escalate across four tiers — Minor → Substantial → Severe → Critical (critical exists only for fatigue) — from ~$5k to ~$20k + demerit points, applied across the chain of responsibility (driver, operator, scheduler).
- Drivers must log work/rest and find compliant rest areas — a large planning burden no consumer nav app addresses. Rest-area availability/occupancy is a known pain point.
2.3 Outback / remote long-distance driving¶
- Fuel gaps: roadhouses 150–250 km apart, often closed weekends/after hours. Running dry is life-threatening.
- Mobile blackspots: reception disappears — kills any always-online app.
- Road trains up to 53.5 m need ~1 km to overtake safely — overtaking guidance matters.
- Wildlife & livestock (kangaroos/cattle at dawn/dusk); weather/hazards (flash floods over creek crossings, dust storms, unsealed/4WD-only roads, corrugations).
- Isolation safety: breakdowns can be fatal; no easy way to raise help.
2.4 Official live-hazard data is fragmented by state¶
Every jurisdiction runs its own feed (NSW Live Traffic, QLDTraffic, VicRoads Data Exchange, Main Roads WA, Traffic SA, plus emergency feeds). No single app aggregates them for a driver — a clean integration win. (Full feed inventory in Part 5.3.)
2.5 Caravans / grey nomads / tourists (adjacent large market)¶
Consumer-scale version of the truck problem: height/length limits, towing routes, dump points, free camps, water, offline maps. >901,000 caravans/campervans (+27% since 2019), 15.2M caravan/camping trips and ~A$14bn spend a year. Shares our infrastructure almost entirely.
Part 3 — Market & opportunity sizing¶
3.1 Numbers (researched)¶
| Segment | Size | Source basis |
|---|---|---|
| Heavy vehicle units (>4.5 t, incl. buses/trailers) | ~1.06M | NHVR Productivity Plan 2025–30 |
| Powered trucks (rigid + articulated) | ~760k (~630k rigid + ~129k artic) | BITRE / HVIA |
| Road-freight businesses | ~64,700 (+4.4% YoY; mostly <5 trucks) | IBISWorld 2026 |
| Road-freight industry value | ~A$74.0bn (2025); volumes +11.5% next 5 yrs | IBISWorld |
| Road-freight workforce | ~183,600; ~28,000 unfilled driver roles | Jobs & Skills Australia |
| Registered caravans + campervans | ~901,000 (+27% since 2019) | caravan industry stats 2024 |
| RV market value | ~A$18.2bn (2025) → ~A$36.8bn by 2035 (7.3% CAGR) | market research |
| ANZ fleet-management market | ~US$0.76bn (2025) → US$1.76bn (2030) | industry reports |
3.2 TAM / SAM / SOM (illustrative, to be firmed in Phase 0)¶
- TAM — all AU drivers who'd benefit from better nav (trucks + RVs + regional/outback motorists) and all fleets needing compliance/telematics. Rough order: consumer app several million potential users; fleet SaaS addressable across ~64.7k businesses / ~760k trucks.
- SAM — realistically reachable v1–v3: professional heavy-vehicle drivers on the eastern seaboard freight corridors + grey nomads doing the "big lap", + small fleets (5–50 trucks). Say ~150–250k professional drivers + ~200k active tourers.
- SOM (3-yr target) — low-single-digit % of SAM: e.g. 30–60k active consumer users and 300–800 fleet vehicles under SaaS. (Firm these with Phase-0 interviews and funnel assumptions.)
3.3 Competitive landscape (researched)¶
| Product | Category | Offline | AU truck data | HVNL compliance | Pricing | Gap it leaves |
|---|---|---|---|---|---|---|
| Google Maps | Consumer nav | Limited | None | None | Free | No truck dims / outback tracks |
| Waze | Social nav | No | None | None | Free (ads) | Useless offline & outback; no truck legality |
| CoPilot Truck | Truck nav | Yes | Decent urban | None | Sub/fleet bundle | Weak AU HVNL/road-train/outback |
| Sygic Truck & Caravan | Truck+RV nav | Yes | AU/NZ region | None | ~€72/yr | Poor outback data; generic |
| SmartTruckRoute AU | Truck nav | Partial | AU-specific | None | Sub | Dated UI, tiny base |
| Hema Explorer Pro | Outback/4WD | Yes (best) | Tracks only | None | A$49.99/yr | No truck-legal road routing / compliance |
| WikiCamps / CamperMate | Camp POIs | Yes | POI only | None | A$9.99 / free | Not navigation, no truck dims |
| Fuel Map / PetrolSpy | Fuel prices | Partial | Fuel only | None | Free | Fuel only |
| Teletrac Navman / EROAD / MTData | Fleet telematics + EWD | n/a | via telematics | Yes (EWD) | ~$35+/veh/mo, hardware, quote-based | Enterprise-only; driver-hostile; no outback nav |
The gap (our thesis): No single product combines AU truck-legal routing + outback/offline coverage + HVNL/EWD compliance + live fuel, and the owner-driver / micro-fleet segment (the bulk of 64.7k businesses) is priced out and underserved by hardware-heavy incumbents. That intersection, app-first and self-serve, is the wedge.
Part 4 — Product: feature set (consumer-nav parity + AU differentiators)¶
4.1 Tier 1 — Consumer-nav parity (table stakes)¶
Turn-by-turn nav (via Navigation SDK — see Part 6), dynamic ETA, live traffic re-routing, crowdsourced reports (crash/hazard/police/camera/closure/traffic) with confirmation & aging, speed-limit + camera alerts, saved places, planned drives, share-ETA/live location, fuel prices, CarPlay/Android Auto, voice-first guidance, gamification, community map corrections.
4.2 Tier 2 — Australian differentiators (why we win)¶
① Heavy-Vehicle Mode ("TruckSafe") - Vehicle profile: type, height, length, width, GVM, axle config, combination (rigid/B-double/road train/PBS level), load (incl. hazmat). - Truck-aware routing: route on a truck-capable engine (HERE/TomTom or self-hosted Valhalla/GraphHopper with truck attributes), then re-rank/penalise against our AU legal-network + restriction layer (gazetted B-double/road-train/HML networks + low-bridge/mass limits). - Low-bridge & clearance strike warnings with audible countdown. - Permit awareness: flag when a route leaves the approved network → hand off to NHVR Portal / Route Planner (save-route deep-link into a permit application); store active permits. - Crowd-verified truck stops, weighbridges, rest bays, loading/unloading notes.
② Fatigue & HVNL Compliance Assistant - Choose ruleset (Standard / BFM-equivalent / ACH); app tracks driving vs rest against the exact HVNL limit tables (Part 8), in the driver's base time zone, 15-min blocks, work rounded up / rest rounded down. - Predictive rest planning: "You must take a 7 h break by 21:40 — here are 3 compliant rest areas on your route with facilities," using the National Formal Rest Areas dataset + state feeds. - Digital work-diary aid + reminders before a breach; export logs. - Explicit disclaimer: an aid, not a legal substitute for an NHVR-approved EWD or written work diary (Part 8.4, Part 10).
③ Outback / Offline Resilience - Offline route + map tiles + hazard cache downloaded before entering blackspots (via the OSM/offline engine — Google tiles can't be cached, Part 6). - Fuel-range planner: given tank range, warn "next reliable fuel in 210 km — last chance to refuel is X"; roadhouse hours & fuel-type availability. - Road-train overtaking assist and wildlife high-risk zone/time alerts (dawn/dusk). - Breakdown/SOS: offline-cached emergency info; nearest help; "share my planned route + check-in" so someone is alerted if you don't arrive; satellite-messenger friendly.
④ Unified Official Hazard Layer - Aggregate NSW Live Traffic, QLDTraffic, VicRoads, Main Roads WA, Traffic SA (+ NT/TAS/ACT), Digital Atlas, HazardWatch (NEMA), BOM into one map layer: floods, fires, closures, roadworks, alpine, restrictions. - Blend official feeds + crowdsourced reports → far better coverage than Waze alone, especially rural.
⑤ Fleet / Operator dashboard (B2B monetisation) - Web portal: assign vehicle profiles, monitor fatigue/EWD compliance, track fleet, receive breach alerts, manage permits, fuel/route optimisation. The revenue engine (Part 12).
4.3 Explicitly out-of-scope for v1¶
Building our own base map/routing graph from scratch; payments/freight brokerage; full ELD hardware; becoming an NHVR-approved EWD (revisit — it's a strong later moat but a heavy certification process). Revisit post-traction.
Part 5 — Data sources & licensing (the crown jewels)¶
This is the single biggest moat and biggest effort. Assemble, normalise and keep-fresh. Licence discipline matters — one non-commercial feed in a paid product is a legal problem.
5.1 Map & routing engines¶
| Component | Option | Notes / licence |
|---|---|---|
| Metro base map + traffic + turn-by-turn | Google Maps Platform Navigation SDK | Enterprise/negotiated; DIY nav banned by ToS; no offline; no truck routing |
| Truck-capable + offline engine | HERE (Truck Attributes + Routing) or TomTom + Geoscape or self-hosted Valhalla/GraphHopper + OSM | Commercial licence (HERE/TomTom) or ODbL (OSM). Provides offline + truck profiles Google lacks |
| Routable AU base graph | Geoscape National Roads (ex-PSMA) | Commercial; good base graph, not HV-attributed by itself |
5.2 Heavy-vehicle legal network & restrictions¶
| Dataset | Authority | Format | Licence | Access |
|---|---|---|---|---|
| National Network Map (unified gazetted networks) | NHVR | Web view only | Not stated | View only; NHVR Spatial API (PROTECTED, account req.) |
| NHVR Spatial Open Data Hub layers | NHVR | GeoJSON/SHP/CSV/KML + WMS/WFS | Unconfirmed/empty | Open download (legacy/VIC-weighted) |
| NHVR Developer Platform (Portal + Spatial APIs; PBS, Access, permits) | NHVR | REST/JSON | Account required | PROTECTED |
| WA RAV / HVS networks (RAV levels, road-train, tandem/tri, oversize, structures) | Main Roads WA | GeoJSON/SHP/CSV/KML + OGC | CC-BY | Open, weekly updates — best open source |
| QLD Heavy Vehicle Routes + Restrictions | DTMR | SHP/TAB/FGDB/KMZ/GPKG | CC-BY 3.0 AU | Open download — static since 5 Apr 2024 (migrated to NNM) |
| NSW HV maps / approved-road lists / IAP | TfNSW | Various | CC-BY 4.0 | Open (transitioned to NNM) |
| VIC HPFV/PBS networks | DTP/VicRoads | GeoJSON/SHP + OGC | CC-BY 4.0 | Open |
Strategy: apply for NHVR Spatial API access for currency; use state CC-BY downloads for geometry substrate; crowdsource local-road low bridges (no authority publishes complete local-road clearance data).
5.3 Low-clearance (low-bridge) data¶
| Dataset | Authority | Format | Licence | Coverage caveat |
|---|---|---|---|---|
| NSW State Roads Vertical Clearances | TfNSW | XLSX | CC-BY 4.0 | State roads only; no regional/local; last upd. Oct 2023 |
| VIC Height Clearance / Truck Curfew maps (~1,150 structures) | DTP/VicRoads | GeoJSON/SHP/CSV/KML | CC-BY 4.0 | Published explicitly for GPS developers |
| QLD/Brisbane Road height restrictions | BCC/DTMR | GeoJSON/CSV + API | CC-BY 3.0 AU | State/council structures |
→ No national clearance dataset. This is the strongest case for crowdsourced user contributions in the whole app.
5.4 Live hazard feeds (researched — see Part 2.4)¶
| Jurisdiction | Feed | Format | Key? | Licence | Real-time? |
|---|---|---|---|---|---|
| NSW | Live Traffic Hazards API (/v1/live/hazards/{category}/all) |
GeoJSON | Free key (60k/day) | CC BY | ✅ (cleanest) |
| QLD | QLDTraffic /v2/events |
GeoJSON | Public key (100/min) or own | CC BY 4.0 AU | ✅ |
| VIC | VicRoads Data Exchange (+ EMV for fire/flood) | JSON/GeoJSON | Email-gated token, 20/60s | CC BY 4.0 / EMV conditions | ✅ (fiddly) |
| WA | Main Roads WebEOC Closures + Incidents | GeoJSON/CSV/KML/SHP | None (open) | CC BY 4.0 | ✅ (5-min) |
| SA metro | Traffic SA ArcGIS MapServer | ArcGIS/GeoJSON | None | CC BY | ✅ (Adelaide only) |
| SA outback | Outback Road Report | HTML page only | — | Gov | ⚠️ No API — scrape/partnership gap |
| NT | Road Report NT (ArcGIS-backed) | ArcGIS REST | None known | NT Gov | ✅ semi (has weight/vehicle restrictions — valuable) |
| TAS / ACT | Service Tas / Roads ACT (planned) | HTML/ArcGIS | None | CC BY | ⚠️ mostly planned |
| National roads | Digital Atlas Roadworks & Closures | ArcGIS/GeoJSON | None | CC BY | ⚠️ daily (backfill only) |
| National emergency | HazardWatch (NEMA) + per-jurisdiction CAP/GeoJSON | GeoJSON/GeoRSS/CAP | Varies | Per-agency (attribution) | ✅ (~60s) — best fire/flood source |
| Weather/flood warnings | Bureau of Meteorology feeds | RSS/XML/CAP | None | ⚠️ "not for commercial use" | ✅ but needs commercial agreement |
Licence flags: clean CC-BY — NSW, QLD, WA, VIC/DataVic, SA metro, ACT, Digital Atlas. Needs agreement — BOM (non-commercial), EMV (must display source + timestamp), VIC Data Exchange (gated). No feed — SA Outback, most of TAS.
5.5 Rest areas, fuel & POIs¶
- Rest areas: National Formal Rest Areas API (National Freight Data Hub) as the spine; TfNSW NSW Rest Areas + Main Roads WA HV Rest Area for facility detail (light/heavy access, toilets, etc.). All open.
- Fuel prices: NSW FuelCheck API (free; NSW + ACT + TAS via V2), QLD Fuel Price Reporting (free), plus VIC/WA/SA/NT feeds; aggregators (Fuel Map, PetrolSpy) as reference. → A live fuel layer is cheaply/freely buildable — strong feature for both sides.
- POIs / places: Google Places (metro; note caching limits — Place IDs indefinite, lat/lng ≤30 days, other fields not warehoused) + OSM + crowdsourced roadhouses/truck stops.
Part 6 — Architecture (revised: dual-engine)¶
The original "everything on Google Maps" design is not ToS-compliant for turn-by-turn and can't do offline or trucks. Revised design:
┌─────────────────────────────────────────────────────────────────────────────┐
│ Mobile apps (iOS + Android) — Flutter or React Native │
│ • Google Navigation SDK (metro turn-by-turn, traffic — Enterprise licence) │
│ • OSM/HERE offline engine (outback + truck routing + cached tiles) │
│ • Routing-mode switch (online-metro ⇄ offline/truck) + profile-aware UI │
│ • Voice-first report UI (Gemini-style), alerts engine, SOS/check-in │
└───────────────┬───────────────────────────────────────────────────────────────┘
│ REST / gRPC + WebSocket (live) offline sync bundles
┌───────────────▼───────────────────────────────────────────────────────────────┐
│ Backend (GCP-native to sit near Google Maps) │
│ • Routing orchestrator → picks Google (online metro) or truck/offline engine, │
│ then RE-RANKS against AU restriction + legal-network layer │
│ • Restriction/HV service (gazetted networks, low bridges, mass/dim, permits) │
│ • Real-time reports service (ingest, geospatial index, confidence/aging, │
│ dedupe) — PostGIS + Redis + Pub/Sub streaming │
│ • Hazard aggregator (pull NSW/QLD/VIC/WA/SA/NT + HazardWatch + Digital Atlas; │
│ normalise to one schema; per-source licence/attribution tags) │
│ • Fatigue engine (HVNL limit tables + ACH; rest-area matcher; base-TZ clock) │
│ • Fuel/roadhouse service (FuelCheck/QLD feeds + crowd) │
│ • Offline bundle builder (route corridor tiles + hazards + fuel + rest areas) │
│ • User / gamification / fleet / moderation services │
└───────────────┬───────────────────────────────────────────────────────────────┘
│
Google Maps Platform + HERE/OSM engine + official AU feeds + our datasets
Google Maps Platform pieces used: Navigation SDK (metro turn-by-turn), Maps SDK (render), Routes API (metro base routing + traffic), Places (search — with caching limits), Roads (snap-to-road), Geocoding.
ToS guardrails baked into design (researched): - No DIY turn-by-turn by stitching Directions + Maps SDK → use Navigation SDK for live guidance. - No caching of Google tiles; Place IDs storable indefinitely, lat/lng ≤30 days, other Place fields not warehoused; routes not persisted to build a dataset. - Don't put Google content on a non-Google basemap → the offline/OSM engine is a separate map surface, not a mix. UI switches engines rather than blending them on one canvas.
Where our IP lives: (1) the AU heavy-vehicle restriction/legal-network dataset + re-ranking layer; (2) crowdsource pipeline; (3) hazard aggregation & normalisation; (4) fatigue engine; (5) offline resilience & bundle builder. Google/HERE give geometry + traffic; we supply the Australian brain.
Key data model (indicative):
- road_segment (id, geom, jurisdiction, sourced_from) → hv_restriction (segment_id, type[height|mass|length|width|hazmat|network], value, network_class[B-double|road-train|HML|PBS-level], effective_dates, source, confidence).
- report (id, geom, type, created_at, confidence, confirmations, expires_at, reporter_rank).
- hazard (id, geom, category, severity, source, source_licence, attribution, valid_from/to).
- rest_area (id, geom, hv_access bool, facilities[], open_hours), fuel_site (id, geom, brand, fuel_types[], price, open_hours, reliability).
- vehicle_profile (height, length, width, gvm, axles, combination, hazmat), driver_fatigue_state (ruleset, base_tz, work/rest ledger).
Part 7 — Truck-routing algorithm approach¶
Google has no native heavy-vehicle model, and we cannot warehouse its route geometry to build our own engine. So:
- Compute candidate route(s) on a truck-capable engine (HERE/TomTom truck profile, or self-hosted Valhalla/GraphHopper with truck costing) using the vehicle profile (height/length/width/mass/hazmat).
- Overlay the AU legal-network layer: verify every segment is on the vehicle's gazetted/notice network (B-double/road-train/HML/PBS). Segments off-network are hard-blocked unless a stored permit covers them.
- Apply restriction penalties/hard filters: low-bridge clearance < profile height + safety margin ⇒ block; mass/length/width limits ⇒ block; hazmat tunnel bans ⇒ block; time-of-day truck curfews ⇒ penalise/block by schedule.
- Blend live hazards + reports (closures, floods, incidents) as dynamic penalties for re-routing.
- Emit warnings with margins and an audible low-bridge countdown; surface "route leaves approved network here → permit needed" with the NHVR Portal hand-off.
- Confidence & disclaimers: where our restriction data is stale/state-roads-only, mark segments "unverified — confirm signage"; never present a Google route as truck-legal.
For metro trips where the vehicle profile fits standard roads, we can use Google Navigation SDK directly; the truck engine kicks in for restricted vehicles / regional / offline. The orchestrator chooses per-trip.
Part 8 — Fatigue engine specification (HVNL)¶
Build on work/rest limits (stable through the 1 Aug 2026 reform); treat Standard/BFM/AFM/ACH as a labelling/transition layer.
8.1 Standard Hours — solo (the core table)¶
| Period | Max work | Min rest |
|---|---|---|
| 5½ h | 5¼ h | 15 continuous min |
| 8 h | 7½ h | 30 min (blocks ≥15 min) |
| 11 h | 10 h | 60 min (blocks ≥15 min) |
| 24 h | 12 h | 7 continuous h stationary |
| 7 days | 72 h | 24 continuous h stationary |
| 14 days | 144 h | 2× night rest + 2× night rest on consecutive days |
Night rest break = 7 continuous h stationary between 10pm–8am (base TZ), or a 24 h stationary break.
8.2 Basic Fatigue Management — solo (extends daily to 14 h)¶
| Period | Max work | Min rest |
|---|---|---|
| 6¼ h | 6 h | 15 min |
| 9 h | 8½ h | 30 min (≥15 blocks) |
| 12 h | 11 h | 60 min (≥15 blocks) |
| 24 h | 14 h | 7 continuous h stationary |
| 7 days | 36 h long/night work cap | — |
| 14 days | 144 h | 24 h stationary + 2× night rest + 2× on consecutive days |
Long/night work = work >12 h in 24 h, plus work between midnight–6am; capped 36 h/7 days. Split-rest rules can reduce the major break to 6 continuous hours.
8.3 AFM / ACH and two-up¶
- AFM: performance/risk-based, no fixed table; negotiated within outer limits (e.g. reset break ≥30 h incl. two night periods). From 1 Aug 2026, delivered as ACH templated Tables of Hours.
- Two-up: second driver rests (incl. approved sleeper berth while moving) — e.g. Standard two-up 24 h allows 5 continuous h rest in-berth-while-moving; separate 52 h / 7-day / 14-day tables apply.
8.4 Work diary & EWD legal hierarchy (critical for positioning)¶
- Written work diary — always valid (required when driving >100 km from base; 28-day in-vehicle copy; duplicate to record-keeper within 21 days).
- NHVR-approved EWD — legally equivalent, can replace the written diary (~12 approved providers; approvals can lapse — check live register).
- Any unapproved app (ours) — NOT a legal substitute. It assists planning/awareness only. → Ship a prominent, persistent disclaimer.
8.5 Engine rules¶
- Calculate in the driver's base time zone, 15-min blocks, work rounded up / rest rounded down; only ≥15 continuous min counts as rest.
- Count sub-24 h periods forward from end of a rest break; 24 h+ periods from end of a major rest break (7 h solo / 5 h two-up).
- Predictive alerts: warn N minutes before a limit; propose compliant rest areas on-route (Part 5.5).
- Don't hard-code penalty dollars — reference the official "Schedule of HVNL Penalties… 1 Aug 2026 onwards" PDF; amounts are indexed and apply across the chain of responsibility.
Part 9 — Offline / outback resilience design¶
Because GMP has no offline, the offline stack is a separate OSM/HERE engine: - Pre-trip bundle builder: for a planned route corridor, package offline map tiles + turn-by-turn graph + cached hazards + fuel sites + rest areas + low-bridge/restriction layer, sized to the route (+ buffer). - Graceful degradation: online metro (Google) → hand to offline engine when signal drops; seamless UI switch, no blended canvas (ToS). - Fuel-range planner: tank range vs next reliable fuel; "last chance to refuel" prompts; roadhouse hours/fuel-type from FuelCheck/QLD + crowd. - Safety layer: wildlife dawn/dusk zones, road-train overtaking assist, SOS/check-in with route-share; satellite-messenger friendliness for no-signal country.
Part 10 — Legal, regulatory & privacy¶
Flag: this is a research summary, not legal advice. AU road/privacy law is largely state-based (8 jurisdictions). This section needs sign-off from an Australian lawyer specialising in transport + privacy + product liability. Items marked [LAWYER CHECK] are the sharpest ones to brief them on.
Speed-camera / police-location alerts — lawful, but design defensively (researched). - GPS-based camera/police-location alerts are legal in all 8 AU jurisdictions. The banned thing is radar/laser detectors & jammers (hardware that senses/interferes with an enforcement signal) — illegal to use, usually to possess, everywhere except the NT. A passive GPS database senses no signal, so it isn't caught. Waze already operates here surfacing police/cameras. - Camera-data availability: fixed / red-light / average-speed (point-to-point) cameras are officially published and always signposted (easy static layer, e.g. Data.NSW, QLD/VIC/SA/WA registries). Mobile/covert cameras are deliberately withheld ("anywhere, anytime") — exactly where lawful crowdsourced real-time reports add value the official data can't. - [LAWYER CHECK] Some radar-detector statutes cover devices that "detect, interfere with, or give warning of" enforcement equipment. The near-universal reading is a pure GPS app isn't caught, but the "give warning of" wording hasn't been comprehensively tested per state — confirm exact wording (NSW/VIC/WA regs). - International contrast → cheap insurance: France (LOPPSI 2) forces generalised "danger zones" instead of exact camera positions; Germany bans operating a ready-to-use camera-warning app while driving. AU has no equivalent ban, but build toggleable per-jurisdiction modes now — a France-style "danger zone" mode and a Germany-style "disable-while-moving" mode — as insurance against policy change. Frame the product as safety awareness, not enforcement evasion.
Privacy (Privacy Act 1988 + APPs) — a moving target, tightening (researched). - Precise location + live-location sharing + crowdsourced reports = handling personal information (location trails are highly re-identifying). Design against APP 1 (privacy policy), APP 3 (collect only what's necessary; OAIC updated Ch.3 guidance May 2026), APP 5 (notify at/before collection, esp. background location), APP 6 (no unexpected secondary use — ad targeting needs care/consent), APP 8 (cross-border accountability if hosting offshore, e.g. Google Cloud regions), APP 11 (security). - Reform status: Privacy and Other Legislation Amendment Act 2024 passed late 2024. A statutory tort for serious invasions of privacy commenced 10 June 2025 — direct litigation exposure for intrusion/misuse, sharpest new risk for live-location sharing. A Children's Online Privacy Code must be registered by 10 Dec 2026 — heightened obligations if under-18s use the app. Further tranches (automated-decision transparency, possible removal of small-business exemption) expected. → Privacy-by-default now; anonymise/aggregate anything shared to a "Waze-for-Cities"-style partner. [LAWYER CHECK] consent flows, retention, offshore hosting.
Driver-distraction / device laws — validates voice-first, with hard rules (researched). - Governed by the Australian Road Rules: Rule 300 (no hand-held phone use; a phone as a nav/driver's aid is allowed only if in a commercially designed, manufactured mount and not touched to operate) and Rule 299 (a visual display unit mustn't show driver-visible content while moving, except driver's-aid navigation). - Passive mounted/CarPlay/Android Auto navigation is lawful for full-licence holders → voice-first is the compliant path (mirrors Waze "Less Chatty" + conversational reporting). - Physically tapping to report/dismiss while moving breaches Rule 300 → design so manual reporting auto-suppresses while moving; rely on voice. L/P1/P2 provisional drivers: no phone use at all in most states → detect/limit and warn. Mobile phone-detection cameras (NSW/QLD/VIC, expanding) actively enforce this. - [LAWYER CHECK] Whether a specific voice command counts as prohibited "use" varies by state/licence class — get a per-state opinion before marketing "report by voice while driving," and hard-restrict provisional drivers.
Product-liability exposure (the serious one) — disclaimers alone won't save us (researched). - A wrong "clear" on a low bridge, or fatigue advice contributing to a breach/crash, is a safety + legal event — and the exposure is materially higher than a passenger nav app because we hold ourselves out as truck-safe. - Australian Consumer Law: digital products carry consumer guarantees (due care and skill, fitness for purpose) that cannot be contractually excluded for consumer supply → a blanket disclaimer will not eliminate ACL liability. Courts (US precedent, persuasive) have refused to let crowdsourced-nav firms "hide behind" use-at-your-own-risk language where negligence/defective-design/foreseeable harm is in play. The test is reasonable care given foreseeable consequences — so "aid, not substitute" framing helps with negligence/reliance but must be paired with genuine reasonable care (accurate, maintained, sourced clearance data). - HVNL Chain of Responsibility: everyone whose decisions affect heavy-vehicle safety (driver, scheduler, operator, consignor — potentially a tool that influences routing/scheduling/fatigue) can bear a primary duty. Wrong fatigue/route advice could draw the app into the CoR chain. - Mitigations to raise with counsel: authoritative sourcing + provenance/currency logging + confidence flags + rapid crowd-correction; mirror NHVR's own "networks are indicative only — confirm with state authority" language; keep fatigue guidance advisory + require driver confirmation; don't present the app as an HVNL system of record; professional indemnity / product-liability insurance; documented safety-testing. [LAWYER CHECK] an AU opinion on (a) negligence/duty for routing errors, (b) non-excludable ACL guarantees, (c) CoR inclusion, (d) disclaimer limits.
Google Maps ToS compliance (Part 6): no DIY nav, caching limits, no Google-content-on-non-Google-basemap. Get the exact clause text for counsel.
Data-licence compliance: honour every feed's licence — BOM is non-commercial (needs an agreement); EMV requires source + timestamp display; attribute all CC-BY sources; confirm NHVR Spatial API terms before commercial reuse.
Part 11 — Cost model¶
11.1 Google Maps Platform (researched; list/PAYG = upper bound, negotiate at scale)¶
Post-March-2025 pricing: per-SKU free monthly allowances (Essentials 10k / Pro 5k / Enterprise 1k), five volume tiers. Navigation SDK is Enterprise/negotiated (~$25→$2 per 1,000 destinations; guidance & reroutes free). Rough order-of-magnitude at list (assumptions: ~4 nav sessions/user/mo, 1 destination lookup each, some map browsing):
| MAU | ≈ GMP list cost / month | Notes |
|---|---|---|
| 10,000 | ~$1.5k–2k | Nav requests dominate |
| 100,000 | ~$15k–18k | Negotiate Enterprise here |
| 1,000,000 | ~$80k–110k | Effective cost typically 40–70% below list at contract |
Levers: use Autocomplete sessions; prefer Places Essentials ($5) over Pro ($17) fields where possible; cap map reloads; push regional/outback traffic to the cheaper offline engine.
11.2 Second engine + data¶
- HERE/TomTom offline+truck licensing — separate line item (per-transaction or per-seat; negotiate). Or self-host Valhalla/GraphHopper on OSM (ODbL) to trade licence cost for infra/eng cost.
- NHVR Spatial API access; state open data is free (CC-BY) but needs ETL/maintenance headcount.
- BOM commercial agreement if we use its warnings.
- Infra: GCP (PostGIS, Redis, Pub/Sub, BigQuery), CDN for offline bundles.
11.3 Cost discipline¶
Model API spend before committing; the dual-engine split is itself a cost lever (offload cheap/free OSM routing in low-value regions, reserve Google for high-value metro). Cache within ToS limits; batch feed pulls.
Part 12 — Monetisation & business model¶
Two-sided. B2B fleet SaaS is the profit engine; B2C drives reach, data density, and funnel.
12.1 Consumer (B2C) — freemium + ads¶
- Free: base nav, crowd alerts, fuel prices, unified hazard layer (ad-supported via Google Ads/PMax inventory — note Waze's self-serve ads are gone; monetise through Google's stack or our own later).
- Pro ~A$9.99/mo or A$59–79/yr (benchmarks: Hema Pro A$49.99/yr, Sygic ~€72/yr): offline outback maps, truck/caravan dimension routing, HVNL rest-area/road-train data, advanced fuel-range planning.
12.2 B2B fleet-compliance SaaS (undercut incumbents; app-first, self-serve)¶
- Starter ~A$15–20/veh/mo: tracking + EWD/fatigue awareness — target the ~64.7k small operators incumbents ignore.
- Pro ~A$30–40/veh/mo: + truck routing/fuel optimisation, mass/dimension compliance, reporting, breach alerts.
- Enterprise: custom, API/integration, video-telematics add-ons.
- Acquisition hook: a free EWD-style logbook aid (à la Netcorp's free EWD) as a funnel → upsell routing/telematics. (Becoming a certified EWD later is a strong moat but a heavy NHVR process — Phase 4+.)
12.3 Why we can undercut¶
Incumbents (~$35+/veh/mo) are hardware-heavy, install-required, quote-based, enterprise-sold. We're software-first (phone as sensor), self-serve, and monetise the consumer side too — so we can price the SaaS to win the long tail of owner-drivers.
12.4 Indicative unit economics (to firm up)¶
- B2C: near-zero marginal cost per free user except GMP calls → keep free-tier users mostly on the cheap offline/OSM engine; monetise Pro conversions + ads.
- B2B: SaaS gross margin high once data pipeline is built; CAC low via owner-driver communities and the free-EWD hook. Model LtV against ~$15–40/veh/mo and fleet retention.
Part 13 — Go-to-market¶
- Beachhead: professional heavy-vehicle drivers on the eastern-seaboard freight corridors (Melbourne–Sydney–Brisbane) where crowdsource density seeds fastest and low-bridge pain is highest. NSW first (best hazard API + most low clearances).
- Second wave: grey-nomad "big lap" tourers (offline/outback + fuel) — shares infrastructure, high word-of-mouth, seasonal.
- B2B entry: small fleets (5–50 trucks) via the free-EWD hook + driver-app pull ("my drivers already use it").
- Channels: owner-driver Facebook groups/forums, truck-stop partnerships, caravan clubs & shows, industry bodies (NatRoad, ATA), state road-authority "Waze-for-Cities"-style data partnerships for credibility + feeds.
- Cold-start fix: seed the map with official feeds first (so it's useful before crowd density arrives), then let crowdsourcing thicken coverage on trucking corridors.
Part 14 — Phased development plan¶
Phase 0 — Validation & foundation (4–6 weeks)¶
- Interview 20–30 truckies / fleet managers / grey nomads; confirm willingness to pay (esp. B2B SaaS price points).
- Secure Google Maps Platform account and start Navigation SDK / Enterprise sales conversation; model API cost at scale; evaluate HERE vs TomTom vs self-hosted Valhalla for the second engine (build a routing spike).
- Apply for NHVR Spatial API access; audit which restriction datasets are open (CC-BY) vs protected vs must-be-crowdsourced.
- Legal review: HVNL "aid not substitute" positioning, camera/police-alert legality, Privacy Act/APP compliance + 2024–26 reforms, GMP ToS clauses, data licences (esp. BOM non-commercial, EMV conditions), product-liability/disclaimer framing.
- Deliverable: validated wedge + tech feasibility spike + data/licence register + legal risk memo.
Phase 1 — MVP: "AuNav for AU" (3–4 months)¶
Navigation SDK turn-by-turn + crowdsourced reports (crash/hazard/police/camera/closure/traffic) with aging, speed-camera alerts, saved places, share-ETA, unified official hazard layer (start NSW + QLD + WA — the clean-API states), live fuel layer. Goal: a usable, sticky consumer nav with better AU hazard + fuel coverage than Waze. Instrument metrics from day one.
Phase 2 — Heavy-Vehicle Mode (3–4 months)¶
Vehicle profiles; truck-aware routing on the second engine with AU legal-network + low-bridge re-ranking; low-bridge/clearance warnings + audible countdown; gazetted B-double/road-train/HML overlay; permit awareness + NHVR Portal hand-off; truck-stop/rest-bay/weighbridge data (crowd + open feeds). Beachhead: professional drivers on freight corridors.
Phase 3 — Fatigue Assistant + Outback resilience (3–4 months)¶
HVNL work/rest tracking (Standard/BFM/ACH) + predictive rest planning (National Formal Rest Areas + state feeds); offline map/route/hazard/fuel bundles; fuel-range planner + roadhouse hours; wildlife/road-train alerts; breakdown/SOS check-in. Ship the "aid not legal diary" disclaimer prominently.
Phase 4 — Fleet B2B + scale (ongoing)¶
Operator web dashboard (profiles, fatigue/EWD compliance, fleet tracking, breach alerts, permit + fuel/route optimisation); expand hazard feeds to all states (incl. NT/TAS + SA outback partnership/scrape + HazardWatch/BOM agreement); caravan/grey-nomad mode; explore becoming a certified EWD; monetisation scale-up.
Cross-cutting from day 1¶
Offline-first + dual-engine architecture; voice-first minimal-tap UX (legal + safety); privacy-by-default; report moderation/anti-abuse; per-source licence/attribution tagging; disclaimer/liability framing baked into UI; analytics/KPIs instrumented.
Part 15 — Team & hiring (indicative)¶
- Founding tech: 1 mobile lead (Flutter/RN + CarPlay/Android Auto), 1 backend/geospatial lead (PostGIS, routing engines, streaming), 1 data engineer (feed ETL, restriction dataset, crowdsource pipeline).
- Domain: a heavy-vehicle/HVNL subject-matter expert (ex-NHVR/industry) — critical for the restriction + fatigue correctness.
- Design: 1 product designer (voice-first, in-vehicle safety).
- GTM/B2B: founder-led fleet sales initially.
- External: transport + privacy + product-liability lawyer (retainer); insurance (professional indemnity).
Part 16 — Risk register¶
| # | Risk | Severity | Mitigation |
|---|---|---|---|
| 1 | GMP ToS blocks intended use / Navigation SDK cost or approval | High | Dual-engine (reduce Google dependence); early Enterprise conversation; legal review of clauses |
| 2 | Restriction-data quality → wrong "clear" on a low bridge (safety+legal) | Critical | Authoritative sourcing + confidence flags + rapid crowd-correction + "indicative only" disclaimers + PI insurance |
| 3 | Cold-start crowdsourcing in low-density areas | High | Seed with official feeds; target freight corridors first |
| 4 | HVNL/legal exposure on fatigue advice | High | "Aid not substitute" + legal sign-off + don't replace approved diary + don't hard-code penalties |
| 5 | Data-licence breach (BOM non-commercial, EMV conditions) | Medium | Licence register; commercial agreements; per-source attribution |
| 6 | Dual-engine complexity + offline sync bugs | Medium | Phase it; strong test harness; clear online/offline handoff |
| 7 | Map/API cost at scale | Medium | Model early; offload to cheap OSM engine; caching within ToS |
| 8 | 1 Aug 2026 HVNL reform (ACH) changes rules mid-build | Medium | Build on stable work/rest limits; ACH as a label layer |
| 9 | Incumbent (Teletrac/EROAD) moves down-market | Medium | Move fast on owner-driver wedge; consumer+B2B two-sided moat |
Part 17 — KPIs / success metrics¶
- Consumer: MAU/DAU, session frequency, report submissions/user, report confirmation rate, retention (D7/D30), Pro conversion %, offline-bundle usage in blackspot regions.
- Data quality: low-bridge coverage vs known structures, crowd-correction latency, hazard-feed freshness/uptime.
- B2B: fleet vehicles under SaaS, per-vehicle ARPU, fatigue-breach-alerts acted on, churn, CAC/LtV.
- Safety (north star): near-miss / bridge-strike-avoided reports; driver-reported "route saved me from X."
Part 18 — Open questions / immediate next actions¶
- Second-engine decision: HERE vs TomTom+Geoscape vs self-hosted Valhalla/GraphHopper — run a routing spike in Phase 0.
- Navigation SDK terms & price: open the Google Enterprise/Mobility conversation now; confirm truck re-ranking + hybrid architecture are permitted.
- NHVR Spatial API access application + licence terms for commercial reuse.
- Legal memo: camera-alert legality, Privacy Act reforms, liability/disclaimers, BOM/EMV licences.
- SA Outback + TAS gaps: partnership or compliant scraping plan (no API today).
- Pin the exact 1 Aug 2026 penalty schedule & work-diary changes into the fatigue engine.
- Firm TAM/SAM/SOM + unit economics from Phase-0 interviews.
Part 19 — One-paragraph pitch¶
A navigation app that feels like Waze — live community alerts, speed cameras, dynamic re-routing — but built for Australia. It runs on Google Maps in the cities and a purpose-built offline/truck engine in the bush, then adds the things every truckie, interstate hauler and outback traveller actually needs: truck-safe routing that avoids low bridges and stays on legal B-double/road-train networks, HVNL fatigue tracking with compliant rest-stop planning, offline maps and fuel-range warnings for the blackspot country, and one unified live layer of every state's flood/fire/closure feed. Consumers get a smarter, Australia-built nav for free; fleets get a compliance dashboard — priced to win the owner-drivers the incumbents ignore — that they'll pay for.
Appendix A — Key source pointers (indicative; verify live before build)¶
- Google Maps Platform: March-2025 pricing & billing docs; Navigation SDK usage/billing (Android/iOS); GMP Terms of Service + Maps Service Specific Terms.
- NHVR: Route Planner / National Network Map (
maps.nhvr.gov.au); NHVR Developer Platform (api-portal.nhvr.gov.au) — Portal + Spatial APIs (PROTECTED); Work & rest requirements; Counting time; Electronic Work Diary; HVA transition FAQs; 1-Aug-2026 penalty schedule PDF. - State HV/clearance open data: Main Roads WA HVS networks (CC-BY, weekly); QLD Heavy Vehicle Routes (static since Apr 2024); TfNSW Vertical Clearances (CC-BY); VicRoads Height Clearance map (CC-BY).
- Hazard feeds: NSW Live Traffic Hazards API; QLDTraffic
/v2/events; VicRoads Data Exchange + EMV; Main Roads WA WebEOC; Traffic SA + Outback Road Report; Road Report NT; Digital Atlas Roadworks & Closures; HazardWatch (NEMA); BOM data feeds (non-commercial). - Rest areas / fuel: National Formal Rest Areas API; TfNSW & Main Roads WA rest-area datasets; NSW FuelCheck API; QLD Fuel Price Reporting.
Prices, licences and endpoints in this document reflect 2026 research and must be re-verified against live official sources (and with counsel) before commitment.