Pattern-of-life analysis

Pattern-of-life analysis builds a baseline of normal, repeated behavior for a subject of interest, a person, vehicle, vessel, or facility, by observing it over an extended period, then watches for departures from that baseline. The underlying assumption is that most activity is routine: the same commute, the same delivery schedule, the same hours of operation, and that a break from routine is often more informative than any single observation on its own.

Analysts typically build the baseline from repeated geospatial intelligence passes, location data, or other time-stamped observations, then apply techniques similar to anomaly detection to flag a deviation for human review rather than automatically treating it as significant. This differs from a single snapshot assessment, which has no history to compare against and cannot distinguish a genuine change from normal day-to-day variation.

The technique is used in defense and security targeting, but also in civilian contexts such as studying Origin-destination (O-D) data for transportation planning. It feeds situational awareness and all-source analysis products. Its central pitfall is that a thin or short observation history produces an unreliable baseline, so an ordinary event can be misread as a deviation, and vice versa.