Database infrastructure for sensor-driven applications
Quasar stores numerical and time-series data and provides query access across incoming observations and retained history.
Model source identifiers, timestamps, calibration context, and revisions alongside the values they describe. This gives applications a queryable record to use for investigation, replay, and subsequent analysis.
The database layer provides:
High-rate and mixed-frequency numerical data
Native timestamps and recorded timing context
Multiple observations and explicitly modeled revisions
Retained history for application replay and analysis
Adaptive, lossless numerical compression
Analytical queries while ingestion continues
Sensor fusion, track correlation, clock correction, and operational interpretation are handled by the applications using this data. Quasar supplies the storage and query infrastructure beneath those workflows.Explore the database architecture →
Compare what was known then with what is known now
See a sensor-fusion application revisit the same event using different knowledge cutoffs. The SQL shown runs against Quasar and uses native historical version selection to retrieve the appropriate stored observations.
Saved replay combining publicly recorded traffic with a fictional sensor scenario. This is a capability demonstration rather than an operational C2 reference.
Watch the demo
Capabilities for sensor and telemetry workloads
Timestamp representation and source-clock accuracy
Native nanosecond-resolution timestamps preserve the represented time values. Match finer-resolution requirements to the supported release and feature availability. Source-clock accuracy and synchronization determine the accuracy of the recorded timing.
Mixed sensor rates in one database
Store platform state, numerical signals, detections, and derived observations with the schema and timing context needed by the application.
Lossless numerical compression
Delta4C adapts to numerical structure while preserving integer and floating-point values bit for bit. Compression results depend on the data.
Replay based on retained observations and revisions
Record revisions, source identifiers, and the relevant timestamps explicitly in the data model. Applications can then retrieve retained observations and selected revisions for analysis. Define the history to retain as part of the deployment design.