VelocityEHS® has announced the publication of three scientific papers on artificial intelligence (AI) and machine learning (ML) in workplace safety and sustainability. These papers, published in the journals Ergonomics; International Journal of Data Warehousing and Mining; and Heliyon, showcase VelocityEHS' innovative in ergonomics risk assessment, ESG data management, and Chemical Safety.

Advancements in Ergonomics Risk Assessment

The first paper, NLP-based Ergonomics MSD Risk Root Cause Analysis and Risk Controls Recommendation, published in Ergonomics and authored by Pulkit Parikh, PhD., Julia Penfield, PhD., Richard Barker, CPE, CSP, Blake McGowan, CPE, and James Richard Mallon, CPE, presents an AI-powered framework that utilizes Natural Language Processing (NLP) to automate the identification of musculoskeletal disorder (MSD) risks and recommend targeted risk controls. 

By leveraging deep learning and expert-driven ML models, this system goes beyond traditional risk scoring to provide actionable insights that improve workplace ergonomics and reduce injuries.

"Traditional ergonomics assessments often stop at producing risk scores, leaving companies without clear guidance on control actions," said Rick Barker, CPE, Senior Director, Solution Strategy." 

Until now, most research using artificial intelligence to combat musculoskeletal disorders has been limited to risk assessment. One of the unanswered questions among researchers is how to enhance the model to offer sustainable improvement strategies.

Julia Penfield, PhD., VP of Research & Machine Learning at VelocityEHS addressed this challenge: “We presented a framework that goes beyond MSD risk scoring.  Along with machine learning, computer vision, and natural language processing can propose risk control recommendations to help organizations achieve their goal to create safer workplaces. To the best of my knowledge, we are the first to take this holistic approach.”

Revolutionizing ESG Data Management

The second paper, Automatic Question Answering from Large ESG Reports, published in International Journal of Data Warehousing and Mining, and co-authored by Pulkit Parikh, PhD., and Julia Penfield, PhD., introduces the first AI-driven system designed to automatically extract and answer questions from extensive Environmental, Social, Governance (ESG) reports. 

ESG reports often exceed 50 pages, making manual extraction for audits, benchmarking, or Scope 3 reporting time-consuming and labor-intensive. Compounding this challenge, audits require answering hundreds of questions, posing difficulties even for experts. Additionally, midsize companies managing Scope 3 reports must manage thousands of suppliers, making it difficult to process ESG data.

"An AI-system could transform this process, enabling organizations to retrieve relevant information and drive informed decision-making effortlessly and efficiently," said Dr. Julia Penfield.

Transforming Chemical Safety with AI-driven SDS Indexing

The third paper, A Machine Learning Driven Automated System to Extract Multiple Information Fields from Safety Data Sheet Documents, published in Heliyon, and authored by Misbah Khan, Julia Penfield, PhD., Aatish Suman, and Stephanie Crowell, presents an AI-powered system designed to automate the extraction of key chemical safety data from Safety Data Sheets (SDS). 

SDS indexing has evolved from storing physical copies to digitally extracting key fields for inventory and risk management. While essential for compliance, manual SDS indexing is labor-intensive, costly and time consuming. An AI-driven solution will automate this process, allowing organizations to access critical chemical information with speed and accuracy.

“Effective chemical data management is essential for workplace safety and regulatory compliance. AI is no longer the future of chemical safety- it’s the present. With automated SDS indexing, we’re setting a new standard for speed, accuracy, and compliance.” says Misbah Khan, Staff Machine Learning Scientist, VelocityEHS. “An AI-driven solution will allow an organization’s team member to quickly retrieve SDS information in case of an accident, improving the response time and potentially saving a life. This blend of innovation and responsibility propels us toward an EHS future that’s both efficient and human centered.”  

The paper concluded that an automated system could improve efficiency and compliance by indexing fields, such as product name, manufacturer, supplier, and revision date, with a precision accuracy of 96 to 99%.

Innovation in Workplace Safety & Sustainability

These research contributions reflect VelocityEHS' commitment to pioneering AI to improve workplace safety and operational performance.

To learn how Velocity’s AI Machine Leaning scientists worked with certified ergonomists to deliver the most comprehensive ergonomics assessment tool, watch this video.

For more about VelocityEHS, visit www.EHS.com.