Work-at-Height Safety Observation Hazard Information Extraction Dataset

#hazard information extraction #named entity recognition #safety control measure extraction #work-at-height safety monitoring #construction hazard identification #occupational safety management

5000
0.2GB
JSON
CATL
MOBIUSI INC

Updated:2026-10-14

AI Analysis & Value Prop

This dataset contains safety inspection and hazard observation records from work-at-height settings, paired with JSON annotations for work activities, hazard sources, exposed persons or entities, locations, and control measures explicitly mentioned in the records. The paired examples map field observations to structured information and can support training and evaluation of models that interpret safety records and extract information in a defined format. Typical uses include work-at-height safety inspections, hazard record organization, and development of structured-output models.

Dataset Insights

Sample Examples

{}

Technical Specifications

Field	Type	Description
file_name	string	File name
location	string	The worksite or location where the hazard is reported to occur.
hazard_source	string	A source of potential accidents or injuries mentioned in the record.
work_activity	string	The work activity or task mentioned in the record.
exposed_objects	array	People or other entities mentioned in the record as potentially affected by a hazard.
control_measures	array	Safety controls, corrective actions, or protective measures explicitly mentioned in the record.
observation_text	string	The original text of a safety inspection or hazard observation.

Compliance Statement

Authorization Type	Proprietary - Commercial AI Training License (No Redistribution)
Commercial Use	Requires exclusive subscription or authorization contract (monthly or per-invocation charging)
Privacy and Anonymization	No PII, no real company names, simulated scenarios follow industry standards
Compliance System	Compliant with China's Data Security Law / EU GDPR / supports enterprise data access logs

Frequently Asked Questions

What does the dataset contain?: It includes original text from work-at-height safety inspections or hazard observations, paired with structured information covering work activities, hazard sources, exposed persons or other entities, hazard locations, and control measures.

What format is used, and which tasks does it support?: The data pairs text with JSON annotations. It can support semantic understanding, information extraction, and structured-output tasks, including supervised fine-tuning and model evaluation. Use the field definitions and annotation guidelines provided with the data as the source of truth.

Are all details fully annotated in every record?: Annotations focus on information explicitly stated in each source record, so the level of detail may vary. If a detail is not present in the text, it should not be assumed to exist. Review the field definitions and examples against the needs of your task.

How can I obtain the data, and what usage rights apply?: Obtain the dataset through its product page or the publisher's designated channel, and refer to the listing for current access details. Permitted uses, commercial rights, and redistribution restrictions depend on the applicable license; confirm the terms before use.

Can't find the data you need?

Post a request and let data providers reach out to you.

Get this Dataset

Verified for Enterprise Use

Cite this Work

@dataset{Mobiusi2026,
  title={Work-at-Height Safety Observation Hazard Information Extraction Dataset},
  author={MOBIUSI INC},
  year={2026},
  url={https://www.mobiusi.com/datasets/0fe370e58928b3561fd6a113e150e0f1},
  urldate={2026-10-10},
  keywords={work-at-height safety dataset, safety hazard information extraction, site observation records, safety inspection data, hazard identification, exposed entities, JSON annotation data, structured output training},
  version={1.0}
}

Using this in research? Please cite us.

placeholder

Products

Scene

Domain

Modality

Task