Preview: new features on DEV

Data visualization

• Rebuilt the charting engine on a unified, declarative model with grouping, aggregation, time-bucketing, and per-series options. (

  #1466

  )

• Shipped the full chart library across three families:

  • basic (area, bar, bubble, pie, etc.) (#1470),

  • statistical (histogram, box/violin, density, SPC control charts) (#1519),

  • and scientific (heatmap, contour, correlation matrix, radar, wind rose, and more) (#1522)

• Added data filters that persist in the URL, usable from both the experiment table and the chart inspector. (#1467, #1468)

Collaboration & privacy

• Experiment join requests: users request to join, admins approve/reject, both sides get emails, plus a revamped collaborators tab. (#1505)

• Per-experiment contributor anonymization across the data table, charts, and exports. (#1469)

Data ingestion

• Large IoT payloads now upload via S3, bypassing the 128 KB MQTT message limit. (#1446)

• Tuned Grafana alerts to eliminate false positives. (#1546)

Mobile

• CMS-managed alert banners (web + mobile) that editors can publish without a deploy. (#1424)

• Device sheet improvements: MultispeQ prioritization + MAC sticker ID (#1530), USB serial (#1547), BT opt-in (#1550).

These new features can be found on the DEV environment and internal testing track of the mobile app. We of course invite anyone to test them in this period and feedback is more than welcome. Of course, do note that there may be (already known) issues & bugs with the features in this stage. In addition to this, a few things are in code review and will land on DEV soon:

• Experiment dashboards (#1527):

  • Build interactive per-experiment dashboards with a visual layout editor and mix-and-match widgets (visualizations, tables, filters, rich text). Pulls the new chart work together into shareable views.

• Generalized data uploads (#1436):

  • Replaces the ambyte-only path with a general pipeline accepting CSV, TSV, Parquet, Excel, JSON, and NDJSON. Land data into user-defined tables, append to existing ones, and track a unified upload history.

• Large IoT into Databricks (#1510):

  • completes the S3 large-payload work by ingesting those payloads into the medallion pipeline via Auto Loader, merging them with the real-time Kinesis stream.

• Mobile MQTT on Cognito credentials (#1548, #1549):

  • lets signed-in users connect to AWS IoT Core over MQTT using their Cognito identity, attaching the IoT policy at credential issuance.

• Branching in the measurement flow (#1511):

  • brings conditional branching workbooks to mobile, so field researchers can run protocols that route down different paths based on answers (previously web-only).

• XLSX export for experiment data (#1371, draft):

  • adds an Excel export option alongside the existing formats.

Finally, as to some release cadence, we are likely looking at:

• a production release of the web platform (and related parts) next week which would include all (working-in-order) features that are present on DEV

• a new mobile app release later today or tomorrow morning to include certain fixes and adjustments to serial communication