What can high-resolution smart-meter consumption data reveal about a household?
Presence (when someone's home or away), activity patterns (cooking, showering, sleep/wake times), which appliances are used (each has a power "fingerprint"), and even household size and changes.
* NILM disaggregates one meter's aggregate draw into presence, activities, appliances and household size. *
Smart meters' high-resolution consumption data allows alarmingly precise inferences about private life:
- Presence detection: exact times someone is home or leaves; holidays and regular absences become visible.
- Activity patterns: when people cook, shower, watch TV; sleep and wake times are detectable; nighttime activity stands out.
- Appliance use: consumption spikes reveal which devices are used — each appliance has a characteristic "fingerprint" in the power draw.
- Household size: the number of occupants can be inferred from consumption patterns; changes (new baby, someone moving out) become visible.
This information interests energy providers, advertisers, insurers, potential burglars, and surveillance authorities — so protecting it is paramount.
Tip: This technique is called NILM (non-intrusive load monitoring) — appliances have distinct power signatures, so a single meter at the fuse box can disaggregate which devices run when, without any sensor inside the home.
Go deeper:
Nonintrusive load monitoring (Wikipedia) — how appliance fingerprints are disaggregated from one meter.- Smart meter — Privacy concerns (Wikipedia) — the privacy implications of high-resolution data.