Published January 14, 2026 | Version v1
Dataset Open

SAFELIFT Dataset: Safety-Aware Feedback for Ergonomic Lifting & Injury-Free Tasks

  • 1. ROR icon University of Bari Aldo Moro
  • 2. ROR icon Marche Polytechnic University
  • 3. ROR icon Delft University of Technology

Description

This dataset supports the research presented in "SAFELIFT: Safety-Aware Feedback for Ergonomic Lifting & Injury-Free Tasks". It consists of annotated videos of individuals performing lifting tasks under controlled conditions. The dataset is designed to facilitate research in computer vision, ergonomics, and human-centered AI, particularly in the development of systems that assess and improve lifting posture using the Revised NIOSH Lifting Equation (RNLE).

Each video captures a person lifting a package from the floor to a target surface, simulating typical workplace lifting scenarios. Annotations include posture-dependent measurements such as:

  • Horizontal hand distance (H)

  • Vertical hand position (V)

  • Vertical displacement (D)

These measurements are aligned with the RNLE and can be used to compute the Recommended Weight Limit (RWL) and the Lifting Index (LI) for ergonomic risk assessment.

Ethics and Privacy:
All participants provided informed consent. The dataset has been approved by TU Delft institutional review board for privacy and ethical compliance.

License:
CC BY-NC-SA 4.0

Suggested Citation:

BibTeX format:
@inproceedings{Dibenedetto26IUI,
  author       = {Gaetano Dibenedetto and Pasquale Lops and Piero Lovreglio and Marco Polignano and Roberto Ravallese and Helma Torkamaan},
  title        = {SAFELIFT: Safety-Aware Feedback for Ergonomic Lifting \& Injury-Free Tasks},
  booktitle    = {Proceedings of the 31th International Conference on Intelligent User Interfaces, {IUI} 2026, Paphos, Cyprus, March 23-26, 2026},
  publisher    = {{ACM}},
  year         = {2026},
  url          = {https://doi.org/10.1145/3742413.3789143},
  doi          = {10.1145/3742413.3789143}
}

ACM Reference Format:

Gaetano Dibenedetto, Pasquale Lops, Piero Lovreglio, Marco Polignano, Roberto Ravallese, and Helma Torkamaan. 2026. SAFELIFT: Safety-Aware Feedback for Ergonomic Lifting & Injury-Free Tasks. In 31st International Conference on Intelligent User Interfaces (IUI ’26), March 23–26, 2026, Paphos, Cyprus. ACM, New York, NY, USA, 16 pages. https://doi.org/10.1145/3742413.3789143

Files

annotations.json

Files (510.9 MB)

Name Size Download all
md5:64d85b200dd175c07f567062db8ab055
179.5 kB Preview Download
md5:872e40d5a6f05c1061088bb66d454245
2.1 kB Preview Download
md5:8dab1a766b208370dbbb6dc96f654813
2.3 kB Download
md5:b64b99fb500339795bd29a88f3170228
25.7 MB Preview Download
md5:45c6297b1429c09c6b59d8a09242b453
25.5 MB Preview Download
md5:8aaeca97486306041eeb0c4ae5d94f5c
4.3 MB Preview Download
md5:72a8eb85cb334e1c7697fbee6e9f41eb
4.4 MB Preview Download
md5:41e00dcadada70dba903647ff66d1949
45.1 MB Preview Download
md5:077fd7284bc6b764d38d341efe1cab5e
51.8 MB Preview Download
md5:a83c7ba3a7f3271429762085a14eb7c9
247.8 MB Preview Download
md5:a0c0e734f180d2d287c4b384a0a25309
106.1 MB Preview Download

Additional details

Related works

Continues
Dataset: https://zenodo.org/records/16743120 (URL)
Publication: 10.1145/3705328.3759314 (DOI)

Software

Repository URL
https://github.com/GaetanoDibenedetto/IUI26
Programming language
Python

References

  • Gaetano Dibenedetto, Pasquale Lops, Piero Lovreglio, Marco Polignano, Roberto Ravallese, and Helma Torkamaan. 2026. SAFELIFT: Safety-Aware Feedback for Ergonomic Lifting & Injury-Free Tasks. In 31st International Conference on Intelligent User Interfaces (IUI '26), March 23–26, 2026, Paphos, Cyprus. ACM, New York, NY, USA, 16 pages. https://doi.org/10.1145/3742413.3789143