# measurement

Node-RED custom node for sensor signal conditioning. Takes raw input — either a single scalar (analog mode) or an MQTT-style JSON object with many keys (digital mode) — and produces scaled, smoothed, outlier-filtered measurements. Part of the [EVOLV](https://gitea.wbd-rd.nl/RnD/EVOLV) wastewater-automation platform.

Registers itself on port 2 as a child of a parent equipment (rotatingMachine, pumpingStation, reactor, etc.). The parent consumes measurements via shared `MeasurementContainer` events.

## Install

```bash
cd ~/.node-red
npm install github:gitea.wbd-rd.nl/RnD/measurement
```

Or pull the whole platform via the superproject. Restart Node-RED and the node appears in the palette under **EVOLV**.

## Two input modes

### Analog mode (default)

One scalar per message — the classic PLC / 4-20mA pattern.

```json
{ "topic": "measurement", "payload": 42 }
```

The node runs one offset → scaling → smoothing → outlier pipeline and emits exactly one MeasurementContainer slot. Every existing flow built before digital mode keeps working unchanged.

### Digital mode (MQTT / IoT)

One object per message, many keys:

```json
{ "topic": "measurement",
  "payload": { "temperature": 22.5, "humidity": 45, "pressure": 1013 } }
```

Each key maps to its own **channel** with independently-configured scaling, smoothing, outlier detection, type, position, unit, and distance. A single inbound message therefore emits N MeasurementContainer slots — one per channel — so a downstream parent sees everything at once.

Pick the mode in the editor or via `msg.mode`. Analog is the default; digital requires populating `channels` (see *Configuration*).

## Input topics

| Topic | Payload | Effect |
|---|---|---|
| `measurement` | analog mode: `number` or numeric `string` — stored as `inputValue` and consumed on the next tick. digital mode: `object` keyed by channel names. | drives the pipeline |
| `simulator` | — | toggles the simulator flag |
| `outlierDetection` | — | toggles outlier detection |
| `calibrate` | — | adjust the scaling offset so current output matches `inputMin` (scaling on) or `absMin` (scaling off). Requires a stable window. |

## Output ports

| Port | Label | Payload |
|---|---|---|
| 0 | `process` | analog: `{mAbs, mPercent, totalMinValue, totalMaxValue, totalMinSmooth, totalMaxSmooth}`. digital: `{channels: {<key>: {mAbs, mPercent, ...}}}`. Delta-compressed — only changed fields emit each tick. |
| 1 | `dbase` | InfluxDB line-protocol telemetry |
| 2 | `parent` | `{topic:"registerChild", payload:<nodeId>, positionVsParent, distance}` emitted once ~180ms after deploy |

## Configuration

### Common (both modes)

- **Asset** (menu): supplier, category, `assetType` (measurement type in the container — `pressure`, `flow`, `temperature`, `power`, or any user-defined type like `humidity`), model, unit.
- **Logger** (menu): log level + enable flag.
- **Position** (menu): `upstream` / `atEquipment` / `downstream` relative to parent; optional distance offset.

### Analog-mode fields

| Field | Purpose |
|---|---|
| `Scaling` (checkbox) | enables linear source→process interpolation |
| `Source Min / Max` | input-side range (e.g. 4–20 mA) |
| `Input Offset` | additive bias applied before scaling |
| `Process Min / Max` | output-side range (e.g. 0–3000 mbar) |
| `Simulator` (checkbox) | internal random-walk source |
| `Smoothing` | one of: `none`, `mean`, `min`, `max`, `sd`, `lowPass`, `highPass`, `weightedMovingAverage`, `bandPass`, `median`, `kalman`, `savitzkyGolay` |
| `Window` | sample count for the smoothing window |

### Digital-mode fields

- **Mode**: set to `digital`.
- **Channels**: JSON array, one entry per channel. Each entry:

```json
{
  "key": "temperature",
  "type": "temperature",
  "position": "atEquipment",
  "unit": "C",
  "scaling":  { "enabled": false, "inputMin": 0, "inputMax": 1, "absMin": -50, "absMax": 150, "offset": 0 },
  "smoothing": { "smoothWindow": 5, "smoothMethod": "mean" },
  "outlierDetection": { "enabled": true, "method": "zScore", "threshold": 3 }
}
```

`scaling`, `smoothing`, `outlierDetection` are optional — the node falls back to the top-level analog-mode equivalents when missing. `key` is the JSON field name inside `msg.payload`; `type` is the MeasurementContainer axis (can be any string — unknown types are accepted).

## State and emit contract

Every channel runs the same pipeline: `outlier → offset → scaling → smoothing → min/max tracking → constrain → emit`. Output is rounded to two decimals. MeasurementContainer events follow the pattern `<type>.<variant>.<position>` all lowercase, e.g. `temperature.measured.atequipment`.

Unknown measurement types (anything not in the container's built-in measureMap — `pressure`, `flow`, `power`, `temperature`, `volume`, `length`, `mass`, `energy`) are accepted without unit compatibility checks. Known types still validate strictly.

## Testing

```bash
cd nodes/measurement
npm test
```

71 tests cover every smoothing method, every outlier strategy, scaling, interpolation, constrain, calibration, stability, simulation, output-percent fallback, per-channel pipelines, digital payload dispatch, registration events, and example-flow shape.

## Production status

Last reviewed **2026-04-13**. See the project memory file `node_measurement.md` for the current verdict, benchmarks, and wishlist.

## License

SEE LICENSE. Author: Rene De Ren, Waterschap Brabantse Delta R&D.