Monitoring is the systematic observation and recording of the state, performance or behaviour of a system over time. It collects measurements, detects deviations from expected conditions, and informs humans or automated controllers so they can act. Monitoring can be continuous or periodic and ranges from simple manual checks to automated, high-frequency telemetry.

Characteristics and common components

Typical monitoring setups include sensors or probes that capture variables; agents or collectors that transmit data; storage for raw and aggregated measurements; analysis engines that compute metrics and identify anomalies; visualization and dashboards for trends and status; and alerting or control interfaces that trigger responses. Important characteristics are sampling rate, accuracy, latency, retention period and granularity.

Types and examples

  • Environmental: air and water quality, weather stations and seismic detectors used for safety and research.
  • Medical: patient vital-sign monitoring, telemetry and wearable sensors for diagnosis and treatment.
  • IT and network: uptime, latency, throughput, application performance, log collection and security event monitoring.
  • Industrial and process: temperature, pressure, vibration and flow monitoring for automation, control and safety systems.
  • Ecological: population surveys, migration tracking and habitat condition monitoring for conservation.

Methods, analysis and alerts

Data analysis methods include thresholding, statistical baselines, trend analysis, anomaly detection and predictive modelling. Aggregation and rollups reduce volume while preserving observability. Alerting strategies must balance sensitivity and noise to avoid false positives and alert fatigue; common tactics include multi-condition rules, suppression windows and escalation policies.

Deployment and lifecycle

Effective monitoring requires planning for data collection, secure transmission, scalable storage, retention policies and regular review. The lifecycle includes instrumentation, calibration, data quality checks, dashboarding, incident response and post-incident analysis. Integration with ticketing and runbooks improves operational response.

Challenges, governance and ethics

Challenges include data volume, signal-to-noise ratio, scalability, latency, and the cost of storage and processing. Monitoring can raise privacy, legal and security concerns, particularly when tracking individuals; responsible deployment requires attention to consent, data minimization, access controls and retention rules. Clear policies and oversight help balance utility with rights and compliance.

Monitoring differs from logging (recording raw events) and from observability (a system property that enables diagnosis of unknown problems). Surveillance is a related term typically applied to human-focused monitoring and carries distinct ethical implications. In many domains, monitoring not only informs decisions but enables automated feedback control to maintain safety, performance and service levels.