Safe human-robot collaboration: A systematic review of risk assessment methods with AI integration and standardization considerations

Auteurs

Morteza Jalali Alenjareghi, Department of Mathematical and Industrial Engineering, Polytechnique Montreal, 2500 Chem. de Polytechnique, Montréal, H3T 1J4, QC, Canada
Samira Keivanpour, Department of Mathematical and Industrial Engineering, Polytechnique Montreal, 2500 Chem. de Polytechnique, Montréal, H3T 1J4, QC, Canada
Yuvin Adnarain Chinniah, Department of Mathematical and Industrial Engineering, Polytechnique Montreal, 2500 Chem. de Polytechnique, Montréal, H3T 1J4, QC, Canada
Sabrina Jocelyn, Institut de recherche Robert-Sauvé en santé et en sécurité du travail (IRSST), Sabrina Jocelyn, 505, De Maisonneuve Blvd. West, Montréal, H3A 3C2, QC, Canada
Abdelhak Oulmane, Department of Mathematical and Industrial Engineering, Polytechnique Montreal, 2500 Chem. de Polytechnique, Montréal, H3T 1J4, QC, Canada

Type de document

Revues de littérature, synthèses de connaissances

Année de publication

2024

Langue

Anglais

Titre de la revue

International Journal of Advanced Manufacturing Technology

Première page

4077

Dernière page

4110

Résumé

Improving risk assessment (RA) for human-robot collaboration (HRC) is crucial, given challenges in existing RA tools. For example, simulation and testing RA tools lack realism due to simplified models, limited dynamic realism, and sensor integration constraints. This study explores how Artificial Intelligence (AI) can enhance risk assessment methods. Using a systematic literature review, the study analyzes three risk assessment methods: PFMEA, HAZOP, and FTA to which AI has been integrated. Results highlight strengths (e.g., systematic process failure identification, thorough hazard identification, complex system modeling) and limitations (e.g., limited coverage of human-robot dynamics, reliance on historical data, adaptability constraints, resource intensity, design dependency, complexity, human factor oversight, data dependency) in addressing HRC risks. Challenges and opportunities of AI in risk assessment are identified, emphasizing reliability, adaptation, safety, and method accuracy. Adhering to standards significantly improves the trustworthiness of AI-driven risk assessments, ensuring consistent, and validated results across diverse HRC scenarios. The study proposes a hybrid approach that combines multiple methods and incorporates image processing as a practical AI feature to enhance risk detection and prevention. The study advances AI-based risk assessment for HRC, offering a comprehensive overview of the current state of the art, highlighting gaps for future research, and suggesting a holistic solution. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

Mots-clés

Robot collaboratif, Collaborative robot, Intelligence artificielle, Artificial intelligence, Évaluation du risque, Hazard evaluation

Numéro de projet IRSST

n/a

Citation recommandée

Alenjareghi, M. J., Keivanpour, S., Chinniah, Y. A., Jocelyn, S. et Oulmane, A. (2024). Safe human-robot collaboration: A systematic review of risk assessment methods with AI integration and standardization considerations. The International Journal of Advanced Manufacturing Technology, 133(9), 4077-4110. https://doi.org/10.1007/s00170-024-13948-3