Self-Organizing-Production-Logistics/MoRoOp

MoRoOp: A Dataset of Autonomous Mobile Robot Operations in Production Logistics

MoRoOp contains 72 hours of operational data from an autonomous mobile robot (AMR). The data was recorded at the KTH IPU Lab as part of the Self-Organizing Production Logistics (SOPL) project. The current release (1.0.0) covers five consecutive days, 2026-07-27 through 2026-07-31, organized into morning and evening shifts.

[image]

Data organization

The data/ directory contains one folder for each of the nine recorded shifts. Every shift folder follows this naming pattern:

data/YYYY_MM_DD_{morning|evening}/
  YYYY_MM_DD_{morning|evening}_jobs.parquet
  YYYY_MM_DD_{morning|evening}_operations.parquet
  YYYY_MM_DD_{morning|evening}_dispatch_events.parquet
  YYYY_MM_DD_{morning|evening}_robot_state_raw.parquet
  YYYY_MM_DD_{morning|evening}_robot_state_cleaned.parquet

data/kits.parquet is a shared kit catalogue with the columns kit_id, color, size, and quantity.

Tables

• —jobs: Job-level records, including payload, operation definitions, lifecycle state and outcome, termination reason, and release, start, due, finish, and creation timestamps.
• —operations: Operation-level records linked to jobs by job_id, including state, outcome, payload, priority, dependencies, retries, termination reason, and timestamps.
• —dispatch_events: Operation Dispatcher event logs, tracing the management of operations on the mobile robot. Includes event type, operation ID, job ID, and timestamps.
• —robot_state_raw: Reindexed, postprocessed but unfiltered robot-state and diagnostic telemetry before postprocessing.
• —robot_state_cleaned: Filtered and cleaned robot-state and diagnostic telemetry. Its IDs are retained from the raw data, so gaps identify removed records.

Code examples

Example code for loading and analyzing MoRoOp, together with scripts for reproducing the publication figures, is available in the MoRoOp dataset toolkit.

How to download the full dataset

The full dataset can be downloaded from the Hugging Face Hub using the following command:

huggingface-cli download --repo-type dataset Self-Organizing-Production-Logistics/MoRoOp --token <your_huggingface_token> --local-dir ./moroop_dataset