Traverse: GIS-Powered Field Operations Platform
We built a spatial platform that put 42,000 government infrastructure assets on a map, with offline-first field inspections, condition heatmaps, and predictive deterioration modelling.
+40%
Inspection Completion Rate
42,000+
Assets Mapped
$2.8M
Emergency Costs Avoided
100%
Offline Capable
The Problem
The agency manages tens of thousands of physical assets spread across a state: bridges, stormwater drains, retaining walls, guardrails, footpaths, signage, culverts, sea walls. Every asset needs periodic inspection. Every inspection needs GPS-stamped photo evidence, condition scoring, and a work order if something needs fixing.
Field crews were using paper forms and spreadsheets. An inspector would drive to a site, fill out a clipboard form, take photos on a personal phone, drive back to the office, and manually enter data into TechnologyOne. Photos were emailed to a shared inbox. GPS coordinates were estimated from Google Maps. None of it was linked.
Basic questions had no answers: bridges in poor condition by LGA, overdue inspections, deterioration trends for coastal assets. The data existed, scattered across filing cabinets, email attachments, and spreadsheet tabs named by month. It wasn't usable.
Why Existing Tools Failed
Government asset management is a solved problem in theory. There are plenty of platforms: TechnologyOne, SAP PM, Assetic, CONFIRM. All designed for the office. They assume reliable connectivity, someone at a desk entering data, and photos uploaded manually.
Mobile coverage in rural Australia is patchy at best. Inspectors work in areas with no signal for hours. A tool that requires connectivity requires paper fallback, which makes it paper with extra steps.
GIS platforms (ArcGIS, QGIS) have the spatial capability but weren't designed for mobile field data capture. Bolting a form onto a map falls short of a field operations platform that happens to be spatial.
The agency needed something that didn't exist: a mobile-first, offline-capable, spatially-aware field operations platform that integrated with their existing asset management system and produced the analytics managers needed.
The Field Inspection Workflow
Everything in Traverse starts with the field crew. A poor inspector experience produces poor data, and poor data makes everything downstream worthless. We designed the workflow from the inspector's perspective first.
Pre-load at Depot
Download map tiles, asset data, and inspection templates for the work area
Offline-first · Vector tiles · Asset cache
Navigate to Asset
GPS-guided routing to next inspection point with asset details pre-loaded
Turn-by-turn · Priority queue · Nearest-first
Inspect & Score
Configurable templates: condition score, GPS lock, timestamped photos, defect flagging
Template per asset type · Photo evidence · GPS auto-lock
Create Work Order
Flag defects, assign priority, auto-generate work order with location and photos
Priority levels · Photo attached · Auto-queued
Sync on Return
All data syncs back to Traverse Cloud when connectivity returns — zero data loss
Conflict resolution · Delta sync · Audit log
Pre-load at Depot
Download map tiles, asset data, and inspection templates for the work area
Navigate to Asset
GPS-guided routing to next inspection point with asset details pre-loaded
Inspect & Score
Configurable templates: condition score, GPS lock, timestamped photos, defect flagging
Create Work Order
Flag defects, assign priority, auto-generate work order with location and photos
Sync on Return
All data syncs back to Traverse Cloud when connectivity returns — zero data loss
Configurable Inspection Templates
A bridge inspection captures different data than a drainage assessment or a guardrail check. Traverse uses configurable templates per asset type, each with its own condition scoring rubric, defect categories, photo requirements, and mandatory fields. Templates are versioned, so when inspection standards change, old inspections remain valid against the template they were captured under.
Photo Evidence System
Photos are first-class objects in Traverse. Each photo is automatically GPS-tagged, timestamped, compass-oriented, and linked to a specific asset and defect. Inspectors annotate photos in-app, drawing circles around cracks, arrows pointing to corrosion. The result is audit-ready evidence, not a folder of unlabelled JPEGs.
Smart Work Orders
When an inspector flags a defect, Traverse creates a work order automatically, pre-populated with location, photos, condition context, and priority. The maintenance team sees it in their queue immediately (or on next sync). No re-keying. No lost paperwork. No three-week delay between “inspector notices pothole” and “maintenance team knows about pothole.”
Route Optimisation
Inspectors often need to visit 20-40 assets in a day. Traverse optimises the route based on asset locations, priority, and road network, minimising drive time and maximising inspections completed. The route adapts if an inspector marks an asset as inaccessible or adds an unplanned stop.
Offline-First Architecture
Offline is the primary operating mode in Traverse. The system works without connectivity and treats sync as a periodic event, not a constant requirement.
At Depot (Online)
Download vector map tiles for work area
Cache asset data + inspection history
Pull latest inspection templates
Sync pending work orders
In Field (Offline)
Full map + asset access from local cache
GPS-stamped inspections stored locally
Photos geotagged + linked to defects
Work orders queued in local DB
Back Online (Sync)
Delta sync — only changed data uploads
Conflict detection on concurrent edits
Photos compressed + uploaded in background
Dashboard updates in real time
Map tiles are stored as vector tiles, compact enough to cache an entire work region on device, detailed enough to see individual asset locations. The app uses Mapbox GL Native for hardware- accelerated rendering with smooth pan/zoom even on mid-range tablets.
Asset data, inspection history, and templates are cached in a local SQLite database. Inspections are written to local storage immediately and queued for sync. On reconnection, only changed records upload (delta sync). If two inspectors somehow inspect the same asset offline, the system detects the conflict and surfaces it for resolution. Both records are preserved, never silently overwritten.
Photos are compressed client-side and uploaded in the background when bandwidth allows. A 40-inspection day with 200+ photos typically syncs completely within 15 minutes on 4G.
Spatial Analytics & Condition Intelligence
The field data feeds a spatial analytics engine that transforms individual inspections into portfolio-level intelligence. Managers can now see the true condition of their asset network and make evidence-based decisions about where to spend.
Bridges
1,240
Drainage
8,400
Roads
24,600
Guardrails
3,200
Signage
4,800
Condition Heatmaps
Heatmaps aggregate individual asset condition scores into spatial zones, showing managers at a glance where the worst clusters are. Filterable by asset type, LGA, inspection date, and condition band. Exportable as PDF for council reports and budget submissions.
Deterioration Modelling
Traverse builds deterioration curves from historical inspection data, tracking how fast individual assets (and asset classes) degrade over time. The model predicts when an asset will cross a condition threshold, enabling proactive maintenance instead of reactive emergency repair.
Risk Scoring
Risk scores combine condition data with consequence factors. A bridge in poor condition on a school bus route scores higher risk than a guardrail on a low-traffic rural road. Risk scores drive inspection scheduling (worst-first) and maintenance budget allocation.
Custom Map Layers
Traverse supports overlay layers: flood zones, heritage areas, land parcels, easements, bushfire risk, soil classification. Managers can ask spatial questions that weren't possible before: assets in the 1-in-100-year flood zone, heritage-listed bridges approaching poor condition.
Integrations
Traverse fills the gap the agency's existing systems couldn't. Asset data flows bidirectionally with TechnologyOne. Map data comes from Mapbox and can import from ArcGIS. Authentication plugs into existing identity providers.
Traverse Platform
Next.js · TypeScript · PostGIS · Mapbox GL
TechnologyOne
Asset MgmtBidirectional asset sync, work order push, financial data
SAP PM
Asset MgmtAsset register sync, maintenance order integration
ArcGIS / QGIS
GISLayer import/export, WMS/WFS services, shapefile support
Mapbox
GISVector tile rendering, offline tile packs, geocoding
Australia Post
APIAddress validation, geocoding for asset locations
Azure AD / Okta
AuthSSO, role-based access for field crews and managers
The Public Portal
An unexpected addition that became one of the most-used features. Traverse includes a public-facing map portal: a read-only view of asset condition data, stripped of operational details, designed for elected officials and the general public.
Local councillors use it to show constituents what's being done about infrastructure in their ward. The agency uses it to justify budget requests with visual evidence. Residents can see that the bridge they complained about is flagged for maintenance, reducing complaint volume and improving trust.
The portal is deliberately separate from the operational system. No sensitive data, no work orders, no inspector names. Just asset locations, condition grades, and planned maintenance status. Transparency without exposure.
Results
Inspection completion rates jumped 40% in the first quarter. Inspectors spent their time inspecting instead of on admin. Data quality improved overnight: GPS-stamped, photo-evidenced, timestamped inspections replaced hand-scrawled forms and best-guess coordinates.
The spatial analytics gave management visibility they'd never had. They could see the true condition of their portfolio on a map and make evidence-based decisions about where to spend maintenance budgets.
The deterioration model identified 23 bridges trending toward critical condition faster than the inspection schedule would have caught them. Early intervention on those 23 bridges avoided an estimated $2.8M in emergency repair costs.
The platform now manages 42,000+ assets across the state. It's become the system of record for infrastructure condition data, not because anyone mandated it, but because it's the only place where the data is reliable, current, and spatial.
| Process | Before Traverse | With Traverse |
|---|---|---|
| Inspection data capture | Paper forms, manual entry | Digital, GPS-stamped, photo-linked |
| Photo management | Personal phones, email inbox | Geotagged, defect-linked, auto-synced |
| Asset condition visibility | Spreadsheet tabs by month | Spatial heatmap, real-time |
| Work order creation | Back at office, re-keyed | In-field, auto-populated |
| Offline capability | None (paper fallback) | Full functionality, delta sync |
| Deterioration forecasting | Not possible | Predictive model from historical data |
Technology Stack
Mobile App
- React Native
- Mapbox GL Native
- SQLite (offline DB)
- Camera + GPS APIs
Backend
- Next.js + TypeScript
- PostGIS (spatial queries)
- pgvector (similarity)
- Redis (real-time sync)
Spatial / GIS
- Mapbox (tiles + geocoding)
- Turf.js (spatial ops)
- WMS/WFS services
- Shapefile import/export
Integrations
- TechnologyOne API
- ArcGIS REST services
- Azure AD (SSO)
- Australia Post (geocoding)