Skip to main content

What's inside:

  • Challenges in managing time-stamped blob data and why a new solution was needed
  • Why time-series databases and object storage fall short for robotics & industrial use cases
  • Benchmarks and comparisons vs. TimescaleDB, MongoDB, and MinIO
  • Key features:
    • FIFO quota to prevent disk overflow on edge
    • Metadata labels for selective replication
    • S3 backend for cloud deployments
    • Unreliable network handling for edge-to-cloud replication

Performance highlights:

  • 16x faster writes vs. TimescaleDB (1MB records)
  • 3x faster reads vs. MinIO (1MB blobs)
  • Significant infrastructure cost savings