Functional and Meta-Functional Engineering Design Problems Dataset
Dataset Summary

The Functional and Meta-Functional Engineering Design Problems Dataset is a curated benchmark consisting of 100 engineering design problems spanning multiple technical domains.

Each problem is represented by a focused product context, a functional design objective, a structured design query, a problem type (functional or meta-functional), and an associated engineering domain.

The dataset is designed to support the evaluation of intelligent agents and large language models in engineering design generation, functional abstraction, and innovation-oriented reasoning tasks.

Supported Tasks

This dataset can be used for:

Engineering design concept generation

Functional reasoning and abstraction

Meta-functional goal-driven innovation

Cross-domain analogy-based design

Design agent benchmarking and evaluation

Dataset Structure

Each sample contains the following fields:

Field	Description
product_focus	Target product or system
design_goal	Functional design objective
design_query	Structured functional abstraction
problem_type	Functional or Meta-functional problem
technical_domain	Engineering domain category
Example
{
  "product_focus": "Railway de-icing device",
  "design_goal": "Remove ice from the rail surface",
  "design_query": "Separation - Material - Solid",
  "problem_type": "Functional problem",
  "technical_domain": "Transportation Engineering"
}

Dataset Size

Total samples: 100

Functional problems: 45

Meta-functional problems: 55

Technical domains: 8

Technical Domains Covered

Transportation Engineering

Energy and Power Engineering

Manufacturing and Industrial Automation

Medical and Healthcare Engineering

Agriculture and Environmental Engineering

Building and Urban Infrastructure

Electronics and Intelligent Systems

Environmental Protection and Resource Engineering

Data Construction Methodology

The design problems were constructed based on engineering design literature, practical product design scenarios, and domain expertise.

The dataset integrates:

Explicit product contexts

Clear functional design objectives

Structured abstraction-based representations

Balanced distribution of functional and meta-functional problems

This design aims to minimize domain bias and improve the robustness of agent evaluation.

Intended Use

The dataset is intended for research and educational purposes, including:

Benchmarking engineering design generation agents

Studying functional abstraction in design reasoning

Evaluating innovation-oriented problem solving

Supporting cross-domain design research

Limitations

The dataset focuses on functional abstraction and does not include detailed engineering constraints or quantitative specifications.

The current version contains 100 problems and may be expanded in future releases.

License

Recommended: CC-BY 4.0

Citation
@dataset{liu_functional_meta_functional_design_problems_2026,
  author = {Liu, Hongwei},
  title = {Functional and Meta-Functional Engineering Design Problems Dataset},
  year = {2026},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/LiuHongwei1992/FunctionalAndMetaFunctionalDesignProblems}
}

Contact

Hongwei Liu
scu_liuhongwei@163.com;liuhongwei@suit-sz.edu.cn