Reproducibility Dataset for AI-Enhanced LCA/TCO in Multimodal Transport

A synthetic benchmark dataset of 14,500 observations supports the reproducibility of an AI-enhanced life cycle assessment and total cost of ownership workflow for sustainable supply chains. It includes variables such as route distances, payload factors, energy consumption, emissions, and cost modifiers. The dataset was created by Tomasz Neumann and last updated on figshare in April 2026.

Use Cases

• Training and validating AI models for predicting greenhouse gas emissions based on operational profiles and scenario data.
• Benchmarking LCA/TCO methodologies using the provided synthetic scenario configurations (S1–S3).
• Analyzing cost-resilience trade-offs in supply chains based on energy prices, carbon pricing, and cost modifiers.
• Evaluating model performance using the reported train-validation-test split and validation metrics (MAE, RMSE, MAPE).

Strengths

• Contains 14,500 observations designed for reproducible AI workflow validation.
• Explicitly reflects the 70/15/15 train-validation-test data structure used in the original study.
• Includes a range of variables covering operational, economic, and environmental dimensions, such as energy carriers and carbon intensities.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is synthetic and representative, not proprietary operational raw data, which may limit real-world generalizability.
• Row count is unknown, which may limit suitability assessment for specific modeling tasks.

Provenance

Source

figshare, authored by Tomasz Neumann.

Collection Method

Synthesized from historical-like operational profiles, generated scenario data, and literature-derived parameter sets.

Time Range

null

Freshness

Last updated 2026-04-27 13:01:03; freshness should be verified.

Geography

null