Water and sewer network modelling, in the cloud.
The cloud platform for water and sewer network modelling.
HydroDSS is a cloud-native modelling platform for water-distribution and sewer networks. Model in the browser, on a shared map, against a server-side hydraulic engine — no desktop, no ArcGIS, no per-seat tokens. Bring your existing model with you — import EPANET .inp or your GIS layers (shapefile, GeoPackage, GeoJSON) and georeference it onto the map, with InfoWorks and InfoWater to follow. A differentiable engine is arriving to calibrate and optimise in an afternoon, and risk and renewals are built in. Every output stays traceable and auditable.
Sister to ZoneDSS and AuditDSS in QuestFeed's infrastructure-intelligence platform.
From your model to a renewals plan.
One path, in one platform — bring your network in, model it in the cloud, and turn the result into prioritised renewals.
Bring your model with you — import from EPANET or the GIS formats you already use, then georeference it onto the map.
One platform for the whole network.
Distribution and sewer modelling in the cloud, model migration and a differentiable engine, risk and renewals built in. It's all arriving with the 2026 launch.
Model the pressurised network.
Cloud, multi-tenant EPANET modelling — pumps, valves, tanks, controls, extended-period simulation. Build, import, edit, and run from the browser against a server-side hydraulic engine. No desktop install, no per-seat licence.
Model the gravity network too.
SWMM 1D modelling for sewer and stormwater, in the same workspace as distribution. One platform for the whole network instead of a separate desktop tool per domain.
Browser-native, multi-tenant, on the map.
A PostGIS asset register as the system of record, a MapLibre GIS view, organisation auth with row-level tenancy. Your model lives in the cloud — versioned, shareable, georeferenced — not in a file on one engineer's laptop.
Bring your existing model with you.
Bring your model from EPANET .inp (lossless round-trip) or straight from GIS — Esri shapefiles, GeoPackage, GeoJSON, KML, and any GDAL vector format become junctions, pipes, and subcatchments. Place a schematic or unknown-CRS network on the real map with two- or three-point georeferencing (Helmert or affine). InfoWorks ICM / WS Pro and InfoWater native models follow.
Calibrate in an afternoon.
A differentiable hydraulic engine, validated against EPANET, is productizing into the platform: gradient-based auto-calibration of roughness, demand, and leakage, plus design optimisation. The slow, manual calibration loop becomes a solve.
Find what fails first.
A configurable likelihood × consequence (LoF × CoF) risk matrix with Business Risk Exposure in dollars — built on a pipe-criticality overlay that runs on the model today — feeding budget-constrained renewals. Every score traceable to its factors.
HydroDSS is in private build ahead of a 2026 launch. The platform is built and running in development; we're making it live over the coming period — so every capability above is marked coming soon.
The engine, the data, and the surfaces you work in.
Five parts give the platform its edge: a differentiable engine that makes calibration a solve, the national-scale data that grounds every model, a cited spec catalog that fills component parameters, a unified register that holds the whole network, and the risk layer that turns it all into a renewals plan.
hydroEngine
A differentiable hydraulic engine, validated against EPANET on the benchmark corpus. It makes auto-calibration and design optimisation gradient-based — and it is productizing into the platform, not yet its production solver.
Explore the engine → The dataFoundation
National-scale cadastral, location, environmental, climate, and regulatory data — fused with live Earth observation (Google Earth Engine, Sentinel-1, ERA5, NISAR). Already in production via ZoneDSS and AuditDSS; the platform is grounded in it.
See the Foundation → The specsSpec catalog
A provenance-tracked catalog mapping nominal pipe terms ("DN225 PVC") to geometry, strength class, and hydraulic roughness — 1,822 cited rows across every major family, plus pumps, valves, and surge devices. Every value carries its source. It auto-populates model parameters so engineers don't hand-key them.
Open the catalog → The registerAsset management
One georeferenced register for water and sewer — the system of record the models run on and the risk layer scores against. Parameters auto-populate from the spec catalog; condition feeds back into the hydraulics.
Explore asset management → The risk layerAsset risk
Likelihood × consequence becomes a criticality index and a defensible renewals plan. The pipe-criticality overlay — the first risk surface — runs on the model today.
Explore asset risk →The engine, the register, and the risk layer are separate concerns. The engine answers what the network is doing physically; asset risk ranks what to renew first. The one piece we hold back is the real-time, self-correcting Bayesian twin — our hydroBAG research line — which stays on the roadmap, held to a research bar, while the modelling platform stands on its own.
Grounded in the data that actually drives risk.
The platform is grounded in named, authoritative data sources — already in production on QuestFeed's stack, already powering decisions in ZoneDSS and AuditDSS.
Live Google Earth Engine pipelines keep the foundation current: ERA5 climate reanalysis, JRC global surface-water dynamics, MODIS evapotranspiration, SMAP soil moisture, BoM gridded precipitation, and active-radar Earth observation (Sentinel-1, ALOS-2, NISAR). Every dataset is used as the agency publishes it — so every value traces back to a named, citable source (BoM, ECMWF, NASA, ESA, JRC, Geoscience Australia, ABS), not an in-house model.
Coverage: NSW · QLD · VIC in production (Australia) · New York City & San Francisco in production (United States). New Zealand, UK, and EU on roadmap; regulatory corpus already global across 21 jurisdictions.
The water vertical of an infrastructure-intelligence platform that already exists.
HydroDSS doesn't start from zero. It inherits the same multi-jurisdictional data engineering and the same governed RDS-backed delivery stack that already powers two production QuestFeed products — applied to a standalone water and sewer modelling platform.
ZoneDSS
Live across NSW, QLD, and VIC, plus New York City and San Francisco. Planning intelligence on 12M+ lots and parcels — complete NSW LEP / SEPP / DCP coverage, QLD state planning policies, Victorian planning schemes, and US zoning for NYC and SF.
AuditDSS
In production. 320 regulatory corpora across 21 jurisdictions modelled as a Bayesian-evidence graph — 13K rules, 147K obligations, 67K conditional-probability tables, 2K enforcement actions.
HydroDSS
Building now, launching 2026. The cloud platform for water and sewer network modelling — distribution and sewer engines, a PostGIS asset register, and a GIS, on the same data spine.
Pricing will be simple and transparent.
We're finalising it now, and we'll publish it plainly when it lands. No quotes to chase to find out what the platform costs.
One platform, the whole water network.
Here is the build order. For transient, floodplain, and groundwater we bring the trusted public-domain solver to the cloud and couple it into one asset and model graph — TSNet, HEC-RAS (USACE), MODFLOW 6 (USGS) — rather than re-inventing the physics. Each item is labelled by where it actually is.
- Coming soon
- In development
- Roadmap
- In research
Transient & surge
via TSNet (method of characteristics)
Surge envelopes and pressure time-histories for valve closure, pump trip, and burst events — built on the trusted TSNet solver, delivered in the cloud.
Earth observation & auto model-build
Google Earth Engine
DEM, watershed delineation, and rainfall-runoff forcing to help build and update models from authoritative data. Earth observation is already in production across QuestFeed's ZoneDSS and AuditDSS; we're extending it to HydroDSS.
Floodplain mapping
HEC-RAS (USACE) coupled to SWMM
Two-dimensional inundation by coupling the gravity network to the trusted HEC-RAS solver. Brought to the cloud and joined to the same asset and model graph.
Groundwater
MODFLOW 6 (USGS)
Standalone groundwater modelling and SWMM↔groundwater coupling on the trusted MODFLOW 6 solver — the subsurface half of the water cycle.
Dam & reservoir monitoring
satellite
Satellite monitoring of reservoirs and basins. Monitoring only — never dam-safety assurance.
Sewer condition AI
CCTV defect detection
Defect detection from inspection video, scored against likelihood and criticality to drive renewals — the condition input to the hydraulic and risk models.
Real-time & SCADA assimilation
research
Assimilating live telemetry to localise leaks and bursts. Held to a research bar — it must beat a plain baseline before it ships. No self-correcting-twin claims.
This is the direction, not a delivery commitment. The differentiable advantage is specific to distribution hydraulics; the wrapped solvers are named, trusted, public-domain tools we bring to the cloud and couple together — not a claim to out-physics the desktop incumbents.
Common questions.
- What is HydroDSS?
- HydroDSS is a cloud platform for water and sewer network modelling. It runs EPANET-parity distribution hydraulics and SWMM sewer and stormwater modelling in the browser, lets you import your existing model and GIS layers, calibrates with a differentiable hydraulic engine, and turns the result into asset risk and renewals. It is in private build, launching 2026.
- Is HydroDSS available yet?
- Not yet. HydroDSS is in private development ahead of a 2026 launch — the platform runs in development today and we are making it live to users over the coming period, so every capability is marked “coming soon.” Its sister products ZoneDSS and AuditDSS are already in production.
- Does HydroDSS replace EPANET or InfoWorks?
- HydroDSS models distribution networks with EPANET 2.3.5 parity and reads your EPANET .inp losslessly, so you can bring an existing EPANET model straight in. It is a migration path, not a fight with the incumbents — InfoWorks ICM / WS Pro and InfoWater native-model import are on the roadmap.
- What can I import into HydroDSS?
- EPANET .inp files (lossless round-trip) and GIS layers — Esri shapefile, GeoPackage, GeoJSON, KML, and any GDAL vector format — which become junctions, pipes, and subcatchments. A schematic or unknown-CRS network can be placed on the real map with two- or three-point georeferencing (Helmert or affine).
- What is the differentiable hydraulic engine?
- hydroEngine is a differentiable hydraulic solver validated against EPANET 2.3.5 on the benchmark corpus. Because it is differentiable end to end, calibration of roughness, demand, and leakage becomes a gradient-based solve rather than manual trial and error. It is productizing into the platform.
- How does HydroDSS handle asset risk?
- Asset risk is a configurable likelihood × consequence (LoF × CoF) matrix scored on your own network. Consequence is read from the native hydraulic model — pipe criticality plus who is affected — so there is no second hydraulic licence to run and reconcile. Risk reads as a banded score and, where consequence is priced, as Business Risk Exposure in dollars, with per-factor explainability, feeding a budget-constrained renewals plan. The pipe-criticality overlay runs on dev today; the full matrix is built and in review. ISO 31000 / AWWA J100 framing.
- Who builds HydroDSS?
- HydroDSS is built by Dr. Ali Ershadi and QuestFeed Pty Ltd, an Australian infrastructure-intelligence company that also runs ZoneDSS (planning intelligence) and AuditDSS (regulatory intelligence) in production.
Build is in flight.
Waitlist and contact channels open later this year. In the meantime, the engine, the Foundation, and the spec catalog are documented in detail — start there.