iot

iot

The iot module provides types and functions for IoT device monitoring, sensor data processing, anomaly detection, and alert management. It is organized into sub-modules for types, anomaly detection, and simulation-based monitoring.

Import

from std::iot::types use {
    SensorReading, DeviceState, Alert, ThresholdConfig, MonitoringConfig,
    sensor_reading, create_alert, threshold_config, symmetric_thresholds, check_thresholds
}
from std::iot::anomaly use { zscore_anomalies, spike_detection, combined_anomaly_detection }
from std::iot::simulation use { monitor_device, monitoring_report }

Core Types

SensorReading

Field	Type	Description
device_id	string	Unique device identifier
timestamp	timestamp	Reading timestamp
value	number	Primary sensor value
unit	string	Unit of measurement
quality	number	Signal quality 0.0-1.0

DeviceState

Tracks device operational state during monitoring:

Field	Type	Description
device_id	string	Device identifier
status	string	"online", "offline", "degraded", or "maintenance"
last_reading	number	Last recorded value
readings_count	number	Total readings processed
anomaly_count	number	Anomalies detected
alert_count	number	Alerts generated

Alert

Field	Type	Description
severity	string	"info", "warning", "critical", or "emergency"
device_id	string	Source device
alert_type	string	"threshold", "anomaly", or "offline"
message	string	Human-readable message
value	number	Value that triggered alert
threshold	number	Threshold that was crossed

ThresholdConfig

Four-level threshold configuration with hysteresis:

Field	Type	Description
high_critical	number	Critical high threshold
high_warning	number	Warning high threshold
low_warning	number	Warning low threshold
low_critical	number	Critical low threshold
deadband	number	Hysteresis to prevent oscillation

Type Constructors

• sensor_reading(device_id, value, unit?, quality?) — Create a sensor reading with current timestamp.
• create_alert(severity, device_id, alert_type, message, value?, threshold?) — Create a generic alert.
• threshold_alert(device_id, value, threshold, is_high?) — Create a threshold alert with auto-severity.
• offline_alert(device_id, last_seen_seconds) — Create an offline alert.
• anomaly_alert(device_id, value, expected, deviation) — Create an anomaly alert.
• threshold_config(high_critical, high_warning, low_warning, low_critical, deadband?) — Create threshold configuration.
• symmetric_thresholds(center, warning_delta, critical_delta, deadband?) — Create symmetric thresholds around a center value.
• check_thresholds(value, config) -> string — Check value against thresholds. Returns "ok", "warning", or "critical".

Anomaly Detection (std::iot::anomaly)

Statistical and pattern-based anomaly detection for sensor data:

Statistical

• zscore_anomalies(series, threshold?) — Z-score based (default threshold: 3.0 standard deviations).
• mad_anomalies(series, threshold?) — Modified Z-score using Median Absolute Deviation (more robust to outliers).
• iqr_anomalies(series, k?) — Interquartile range based (default k: 1.5).

Rolling Window

• rolling_zscore_anomalies(series, window?, threshold?) — Adaptive Z-score in a sliding window.
• spike_detection(series, window?, spike_threshold?) — Detect sudden spikes/drops relative to recent baseline.

Pattern-Based

• flatline_detection(series, window?, epsilon?) — Detect stuck sensors (no variance).
• oscillation_detection(series, window?, threshold?, min_oscillations?) — Detect rapid oscillation patterns.
• drift_detection(series, baseline_window?, drift_threshold?) — Detect gradual drift from baseline.

Contextual

• time_aware_anomalies(readings, day_thresholds, night_thresholds, day_start_hour?, day_end_hour?) — Different thresholds for day/night periods.

Aggregation

• combined_anomaly_detection(series) — Flags anomaly if any method detects one (Z-score OR IQR OR spike).
• consensus_anomalies(series, min_agreement?) — Only flags if multiple methods agree.
• count_anomalies(anomaly_flags) -> int — Count true values.
• get_anomaly_indices(anomaly_flags) -> Array<int> — Get indices of anomalies.

Device Monitoring (std::iot::simulation)

Simulation-based device monitoring using the core simulation engine:

monitor_device(readings, config)

Monitor a single device’s sensor readings. Uses Welford’s online algorithm for running statistics and Z-score anomaly detection.

let result = monitor_device(temperature_readings, {
    device_id: "sensor-001",
    thresholds: symmetric_thresholds(25.0, 5.0, 10.0),
    offline_timeout: 300,
    anomaly_threshold: 3.0,
    collect_alerts: true
})

monitor_sensors(sensors, config, correlation_check?)

Monitor multiple correlated sensors with optional cross-sensor correlation checks.

monitoring_report(result)

Generate a summary report from monitoring results.

See Also

• Simulation — core simulation framework