Change the source. Keep the history.

Public robot datasets in a simulated transport workflow. Replay time is relative, not a shared real world timeline.

Replay details

Workflow WF-TRANSPORT-001, simulated. A and B use relative replay windows that start at 0 s. Simulated source C begins after B's final event.

Sources

Each source feeds the same workflow history.
Source	Format	Events in history	State
A: MoRoOp, KTH REAL public dataset Details Klein, Jan-Felix (creator); Jeong, Flores-García, Khilji, Wiktorsson, Baalsrud Hauge (contributors). MoRoOp: A Dataset of Autonomous Mobile Robot Operations in Production Logistics, v1.0.0. KTH Data Repository. https://doi.org/10.71775/kth.0qdea-h5385 License: CC BY 4.0 Assigned replay interval: From 0.000000 s, open created_at stored in Parquet as a UTC adjusted timestamp. Replay 0 s is the window start.	Parquet	13 13 dispatch events	Connected
B: DEDICAT 6G, WINGS ICT REAL public dataset Details WINGS ICT Solutions (Greece). DEDICAT 6G Smart Warehousing robots' datasets. Zenodo. https://doi.org/10.5281/zenodo.10793054 License: CC BY 4.0 Assigned replay interval: 0.000000 s to 164.352000 s Datetime has no documented timezone and is kept as text. Replay 0 s is the first row of the file.	CSV	25 19 robot actions, 6 technical services	Disconnected. History retained.
C: Synthetic work orders · SIMULATED Details Synthetic data generated for this demo by generators/c_generate.py (seed 20261025). License: Generated for this demo Assigned replay interval: From 164.352000 s, open ts_ms is epoch ms on a simulated clock. The first event is placed at the connection start point.	JSON Lines	24 24 work orders	Connected

B disconnected. C connected. B's 25 events remain queryable.

Around the swap

B: original timezone unknown. Replay times shown to 0.01 s, and in full on the two rows that meet at the swap. Full precision is in each original record.

Showing 8 selected rows around the swap.
Replay time (s)	Source	Category	Event
149.06	B · disconnected	Robot action	Release trolley
149.12	B · disconnected	Technical service	Rosping service
152.68	B · disconnected	Robot action	Reverse robot
164.351650	B · disconnected	Robot action	Move robot
Source changed: B to C, at replay 164.352 s
164.352000	C · SIMULATED Added after the swap	Work order	Work order completed
169.35	C · SIMULATED Added after the swap	Work order	Work order completed
174.35	C · SIMULATED Added after the swap	Work order	Work order completed
179.35	C · SIMULATED Added after the swap	Work order	Work order completed

Export your history. Query it outside Base1.

Includes original files, provenance and the same query.

Technical detail

The export is a ZIP with the original files, the sources, raw_files, events and assignments tables as Parquet, the exact SQL, the replay configuration, provenance, attribution, a README and a SHA-256 manifest.

The independent reader uses only the Python standard library and DuckDB. It checks the manifest, runs the same SQL on the export, compares the result with the API and compares the retained history with records saved before the swap.

Inactive in this preview. Download the recorded export below and run the same check yourself.

base1-export-20261025T185148212796Z.zip, 16 files, 14.8 MB.

Independent export verification

B history 25/25 unchanged

Export vs API: Identical
Files: 15 checksums match
Result: PASS

Show full verification output

$ python verify_export.py base1-export-20261025T185148212796Z.zip --api-result base1-export-20261025T185148212796Z.api-result.json --baseline b-events.json
Manifest            15 files, 0 mismatched, 0 unlisted
Original files      4 recorded, 0 mismatched
Same SQL, export    62 events for WF-TRANSPORT-001
Export vs API       62 vs 62 events, identical
Baseline vs export  25/25 events of dedicat6g_wings unchanged
Baseline vs api     25/25 events of dedicat6g_wings unchanged
Result              PASS

Repeat the verification yourself

Download the export from this recorded run and check it outside Base1. You need Python 3 and DuckDB (pip install duckdb).

Recorded export, ZIP with original files, tables, query, provenance and attribution
verify_export.py, the independent reader (also inside the ZIP)
API result captured from the app at export time
B baseline, B's events saved before the swap

Save all four files in the same folder, then run:

python verify_export.py base1-export-recorded-run.zip --api-result api-result.json --baseline b-events.json

Expected result: PASS, with 25/25 B events unchanged.

Measured results of the attempts, with commits, source versions and replay config: results.json, summary in results_summary.txt.

Technical details

Events per source and category

Counts per source and category, from the same query. Categories are not added together.
Source	Category	Events
A: MoRoOp, KTH	Dispatch event	13
B: DEDICAT 6G, WINGS ICT, disconnected	Robot action	19
B: DEDICAT 6G, WINGS ICT, disconnected	Technical service	6
C: Synthetic work orders	Work order	24