Atlas Water Network Optimization Platform

A
Atlas
Water Optimization
Demo
Atlas Engine v2.4.1 — Ready
System Pressure
48.3 psi
▲ 1.2% vs yesterday
Energy Cost / Day
$2,847
▼ 6.4% TOU optimized
NRW Rate
18.3%
→ Threshold 20%
Pump Stations
7/8
⚠ P3 maintenance
🗺 Zone Status
Click any card to view its timeseries data
📉 24h Demand + PressureRiverdale Metro
⚡ TOU Energy Cost
📋 Recent Activity
🔁 Recent Runs
Run IDExampleSolverObj ValueItersDurationStatusAction
EPANET Riverdale Metro
🗺
Click any network
node to inspect
🔗 Modelica / Domain Component Registry
IDTypeClassPortsParametersEquationsStatus
📂 Data Files
⊕ Import New Data
📥
Drop files or click to browse
.npy · .csv · .json · .xml · .inp (EPANET)
👁 Preview — Select a file
MetricValue
Select a data file
⚗️ Preprocessing Pipeline
🔄 Workflow Runner — Riverdale Metro
📋 Step Detail —
Click a step or run the workflow.
📊 Step Metrics
Run a workflow to see metrics.
🖥 Execution Log
⚙️ OCP Configuration
📉 Solver Progress
Idle
Configure and press Solve to start.
Iteration
Objective
‖∇f‖
CPU (s)
📊 State Trajectory
🎛 Control Schedule
Total Energy
3,421 kWh
▼ 8.2% vs baseline
Cost Saved
$412
▲ TOU shift
Min Pressure
32.1 psi
→ Min 28 psi
Pressure Deficit
0 nodes
✓ All satisfied
📉 Demand + Pressure (EPS)
⚡ Pump Energy by Hour
🍩 Water Balance
📋 Pump Schedule — Optimized vs Baseline
🧮 Zone Hydraulic Summary
📊 Scenario Comparison
MetricRiverdale MetroZone 5 MPCReservoir AllocCasADi OCPHydro SystemRaven Hydrology
🖥 System Console
🚀 Atlas Platform Hub
9 engineering frameworks · each with a quantified demo solution · powered by CasADi, RTC-Tools, Raven & Modelica
🎯
New here? Start with the Solution Brief
For every module, get the Idea → Challenge → Solution in one screen — quantified outcomes, the physics engines under the hood, data lineage and live twin status.
💧
Atlas Water Networks
Demo solution — Riverdale Metro Water District
Urban water distribution OCP integrating EPANET hydraulic simulation with CasADi NLP pump scheduling, zone-level MPC real-time control, and NRW diagnosis across networks up to 100k nodes.
Pump scheduling MILP / NLP EPANET 100K-node hydraulics Pressure zone MPC NRW analysis
Active EPANET CasADi / IPOPT RTC
🏔
Atlas Reservoir Management
Demo solution — Multi-District Reservoir Allocation
Multi-reservoir LP yield optimisation with level-pool routing, inter-district water trading, Tennant environmental flow compliance, FEWS PI-XML data exchange, and 30-year planning horizons.
LP yield optimisation Level-pool routing E-flow compliance Inter-district allocation
Active LP / MILP FEWS PI-XML Yield Opt
Atlas Hydropower Dispatch
Demo solution — Multi-Unit Hydro System + Pumped Storage
Multi-unit hydro OCP with RTC-Tools goal programming, pumped-storage TOU arbitrage, unit commitment MILP, reservoir level routing, and Tennant e-flow constraints over 48-hour rolling horizons.
RTC-Tools dispatch OCP Pumped storage arbitrage Unit commitment MILP E-flow / Tennant method
Active RTC-Tools Modelica CasADi / IPOPT
🌧
Atlas Hydrology & Flood
Demo solution — 7-Catchment Basin, HBV + HEC-RAS 2D Flood Routing
Semi-distributed rainfall-runoff using Raven HBV/GR4J per HRU, HEC-RAS 1D/2D flood routing across 18,500-cell domains, 30-year RCP4.5/8.5 climate scenarios, and SWMM urban drainage.
Raven HBV rainfall-runoff HEC-RAS 1D/2D routing 30-yr climate scenarios Flood inundation mapping
Active Raven HBV HEC-RAS 2D SWMM
Atlas Mine Water Balance
Demo solution — Gold Mine TSF + Underground Dewatering OCP
Mine water balance OCP with CasADi NLP — TSF freeboard control to ANCOLD guidelines, return water recycling, underground dewatering TOU pump scheduling, heap leach optimisation, and TDS evaporation tracking.
TSF freeboard OCP (ANCOLD) UG dewatering scheduling TOU pump cost minimisation Heap leach water balance
Active CasADi / NLP ANCOLD GoldSim-type
🌊
Atlas Polder / Canal RTC
Demo solution — Woudagemaal Drainage System (Friesland, NL)
Dutch-style lowland polder real-time control using lexicographic GoalProgramming OCP and Modelica hydraulic components — P1 flood safety → P2 boezem service level → P3 TOU pump energy cost.
Lexicographic goal programming Mixed-integer polder control Tidal boundary OCP Modelica acausal components
Active GoalProgramming Modelica RTC-Tools
🏭
Atlas Wastewater Treatment
Demo solution — Ashbridges Bay WWTP, Toronto
Batch MILP scheduling under Ontario TOU tariffs. Shifts aeration load to off-peak, saves 13% energy without changing treatment process.
Live demo MILP Scheduling TOU optimisation
🏥
Atlas Asset Health
Condition monitoring, predictive maintenance and anomaly detection for pumps, turbines, dam embankments and pipe networks using ML and physics-based models.
SCADA anomaly detection Pump curve degradation Predictive maintenance
Coming Soon ML / Physics SCADA
🤖
Atlas Intelligence
Embedded AI analyst for scenario generation, anomaly explanation and natural-language query across all Atlas framework domains — not a chatbot, a domain-aware engineering assistant.
Natural-language query Scenario generation Cross-domain synthesis
Coming Soon LLM Scenario Gen
⚙️ Atlas Core Engine — System Status
CasADi / IPOPT NLP Solver
Available · v3.6.3
SciPy SLSQP (Linear fallback)
Available · v1.11
NumPy Rule-Based (backup)
Available
FastAPI REST endpoint
localhost:8000 · OK
Raven HBV Hydrological Model
7 catchments active
HEC-RAS 2D Solver
18,500 cells · warm start
EPANET Hydraulic Simulator
100K nodes · steady-state
Atlas Asset Health (ML)
Not installed
🔭 Capability Roadmap — Specified & Ready to Build
🌊 Flood Control vs. Revenue Pareto
Spring freshet trade-off: epsilon-constraint LP sweep produces a real Pareto frontier. Tells clients exactly how much revenue they give up per unit of flood margin — "how much revenue for how much safety."
Pareto LP 2 objectives ✓ Design-ready · 2–3 days
🔧 Maintenance Window Scheduling
Extends the unit-commitment MILP: binary mu,w selects the cheapest maintenance week per unit from eligible windows. Forces unit offline only during its chosen week — minimises lost revenue from downtime.
MILP binary unit commitment ✓ Design-ready · 1–2 days
📋 Recent Scenarios
📊 Data Loaded
Water network files
12 files
Hydro time series
8,760 h/yr · 3 units
Raven hydrology data
30yr · 7 catchments
Mine water balance
168h OCP horizon
🎯 Solution Brief
The one-screen pitch for each digital twin — idea, challenge, solution, proof, and what's under the hood.
💧 Water Networks | Solver: Idle | Zones: 8 active | Data: 292 MB loaded |
🚀 Atlas Platform Hub
Launch modules · monitor system health · access all Atlas frameworks
💧
Atlas Water
Urban water distribution OCP — pump scheduling, pressure management, NRW analysis. Integrates EPANET/SWMM hydraulics with CasADi NLP and MILP solvers.
Active EPANET CasADi RTC
Atlas Hydropower
Multi-unit hydro dispatch OCP with pumped storage arbitrage, TOU market integration, reservoir level routing, and environmental flow compliance.
Active Dispatch OCP