Build a Desalination Plant Reverse Osmosis Membrane Cleaning Scheduler — implementation

## Build
Monitor reverse osmosis membrane pressure differentials and permeate quality to schedule clean-in-place operations. Optimize cleaning chemical usage while maintaining continuous baseline potable water production across the plant.

## User
Desalination Plant Maintenance Manager

## Workflow
1. Ingest SWRO membrane pressure drop telemetry and permeate conductivity readings.
2. Evaluate membrane fouling rates using WaterTAP module performance models.
3. Schedule staggered clean-in-place cycles across multiple RO trains to avoid complete shutdown.
4. Generate a chemical dosing and train isolation manifest for operators.

## Interface
Maintenance scheduling dashboard highlighting membrane fouling trends and recommended chemical wash timelines.

## Data
Utilize WaterTAP RO flowsheet modeling data to generate synthetic membrane fouling degradation profiles.
Starter examples below are synthetic. They are not downloaded records.

## Source
[WaterTAP treatment modeling](https://raw.githubusercontent.com/watertap-org/watertap/main/README.md)
Applies WaterTAP membrane unit models to correlate pressure differential increases with expected fouling severity.

## Tools
agno, pandas, fastapi.
Agno provides agentic reasoning capabilities to evaluate multi-train maintenance schedules against continuous production constraints.
Local or cloud LLM capable of evaluating tabular telemetry logic and generating valid JSON maintenance manifests.

## Deliverables
Submit runnable source, pinned dependencies, setup commands, a working interface, example outputs, and a recorded end-to-end demo. Include tests and an action log.

## Acceptance
- Calculates membrane fouling progression from differential pressure telemetry.
- Determines the optimal clean-in-place cycle timing based on a pressure threshold limit.
- Ensures scheduled RO train outages maintain plant production constraints.
- Generates a human-reviewable chemical cleaning dispatch schedule.

## Starter Examples
```json
[
  {
    "input": {
      "train_id": "RO-TRAIN-A",
      "delta_p_bar": 2.8,
      "baseline_delta_p_bar": 1.5,
      "plant_demand_m3h": 5000
    },
    "expected_output": {
      "fouling_status": "SEVERE",
      "cip_required": true,
      "pressure_diff_bar": 1.3,
      "chemical_dosage_liters": 130,
      "_note": "Simplified synthetic app record; chemical_dosage_liters = pressure_diff_bar * 100"
    }
  },
  {
    "input": {
      "train_id": "RO-TRAIN-B",
      "delta_p_bar": 1.7,
      "baseline_delta_p_bar": 1.5,
      "plant_demand_m3h": 5000
    },
    "expected_output": {
      "fouling_status": "NORMAL",
      "cip_required": false,
      "pressure_diff_bar": 0.2,
      "chemical_dosage_liters": 0,
      "_note": "Simplified synthetic app record"
    }
  }
]
```

## Safety
Maintenance schedules are advisory and require human review by a certified chemical engineer before executing physical clean-in-place operations.
Use mock external writes. Require human review before any real deployment.