What it is
Privacy-preserving sensing redesigns collection so a service receives the minimum information it needs. Techniques can include edge counting, short retention, aggregation, access logging, and independent privacy review; none eliminates the need to justify collection.
How the system fits together
Physical and digital layer
Document the equipment, software, interfaces, and geographic scope as separate maintained components.
Data and identity
Record data owners, timestamps, identifiers, quality limits, access rules, and retention before combining feeds.
Operating workflow
Name the team that reviews outputs, can override automation, and remains responsible during faults or outages.
Evaluation
Compare service, equity, safety, reliability, and cost against a published baseline rather than counting installed devices.
- 01
Physical and digital layer
- 02
Data and identity
- 03
Operating workflow
- 04
Evaluation
Where cities use it
- Pedestrian and vehicle counts
- Environmental observation
- Bounded public-space studies
Data generated
- Aggregated counts
- Device audit logs
- Deletion and access records
Analytics and AI applications
- Computer vision
- Classification
- Privacy-risk assessment
Example in practice
A street study can retain hourly counts while discarding source imagery at the edge, then publish error rates and a fixed end date.
Design considerations
Choose sensor placement and refresh rates from an operational need. Keep device identity, calibration, network health, and location visible to the team using the data.
Limits and risks
- Integration can concentrate security and vendor dependencies.
- Uneven coverage or stale records can make a precise interface misleading.
- A technical capability does not establish necessity, public value, or a measured outcome.