ISOMORPH: Multivariate Logistics Time-Series from a Supply Chain Digital Twin

Isomorph2026 released the full ISOMORPH dataset on 2026-05-19. It contains 49 rollouts of multivariate logistics time-series data generated by a digital-twin simulator of a multi-echelon supply chain with an explicit routing network. All runs use a fixed seed and a horizon of 52,560 time steps.

Use Cases

• Benchmarking supply chain forecasting models based on multivariate logistics time-series.
• Analyzing the impact of routing network dynamics based on the Markov chain formulation mentioned in the description.
• Evaluating model robustness based on the 20 Latin-hypercube uncertainty quantification perturbations.
• Studying multi-echelon inventory and logistics behavior based on data from the digital-twin simulator.

Strengths

• Dataset consists of 49 distinct simulation rollouts.
• Includes 27 mixture scenarios and 20 Latin-hypercube uncertainty quantification perturbations.
• Data is generated from a controlled simulation with a fixed seed (2025) and horizon (T=52,560), ensuring reproducibility.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and data size are unknown, which may limit suitability assessment.
• Data is synthetic, generated by a simulator, and may not fully reflect real-world complexities.

Provenance

Source

Isomorph2026 via Hugging Face.

Collection Method

Generated by a digital-twin simulator of a multi-echelon supply chain with an explicit routing network.

Freshness

Last updated 2026-05-19 20:16:23; freshness should be verified.