
Sign up to save your podcasts
Or


About the Guest
Dr. Elif Erkal is a construction safety research expert who specializes in predicting serious injuries and fatalities (SIF) using data analytics. She currently serves as the Associate Director of Research and Strategy at the Construction Safety Research Alliance (CSRA) at CU Boulder. Her groundbreaking work challenges traditional safety metrics and advocates for the use of predictive models in high-risk environments.
Key Takeaways:
📉 The TRIR Illusion
Traditional metrics like TRIR are not just flawed; they are statistically invalid when it comes to predicting fatalities. Instead of merely counting injuries, organizations need to identify the real "precursors" that signal a major incident is coming.
🤖 AI and Predictive Analytics
Companies are sitting on mountains of unstructured safety data, from inspection reports to near-miss cards. By utilizing machine learning, organizations can shift from simply "collecting data" to actively predicting SIF exposure. However, algorithms must be carefully managed to ensure human safety managers don't become complacent.
⚡ High-Energy Controls
Not all safety controls are created equal. Instead of relying on administrative controls like rules and signs—which are known to fail—organizations should focus on High-Energy Control Assessments (HECA) to implement direct, effective controls.
đź§ą Eliminating "Safety Clutter"
When looking at Environmental, Social, and Governance (ESG) frameworks, true safety culture goes beyond simply listing injury rates. Sometimes, removing outdated rules—known as "Safety Clutter"—is just as critical as creating new ones. Ultimately, safety is about stopping the practice of counting failure and starting to measure capacity.
📚 Resources & Contact
To learn more about Dr. Erkal’s research on predictive safety, check out the following resources:
* Website: CSRA*
* LinkedIn: Elif Erkal, PhD
*All of their research and resources are publicly available and free to access, including their published papers. Since the work is publicly funded, listeners are able to explore the papers, literature, and additional resources available on their website.
By Jowanza JosephAbout the Guest
Dr. Elif Erkal is a construction safety research expert who specializes in predicting serious injuries and fatalities (SIF) using data analytics. She currently serves as the Associate Director of Research and Strategy at the Construction Safety Research Alliance (CSRA) at CU Boulder. Her groundbreaking work challenges traditional safety metrics and advocates for the use of predictive models in high-risk environments.
Key Takeaways:
📉 The TRIR Illusion
Traditional metrics like TRIR are not just flawed; they are statistically invalid when it comes to predicting fatalities. Instead of merely counting injuries, organizations need to identify the real "precursors" that signal a major incident is coming.
🤖 AI and Predictive Analytics
Companies are sitting on mountains of unstructured safety data, from inspection reports to near-miss cards. By utilizing machine learning, organizations can shift from simply "collecting data" to actively predicting SIF exposure. However, algorithms must be carefully managed to ensure human safety managers don't become complacent.
⚡ High-Energy Controls
Not all safety controls are created equal. Instead of relying on administrative controls like rules and signs—which are known to fail—organizations should focus on High-Energy Control Assessments (HECA) to implement direct, effective controls.
đź§ą Eliminating "Safety Clutter"
When looking at Environmental, Social, and Governance (ESG) frameworks, true safety culture goes beyond simply listing injury rates. Sometimes, removing outdated rules—known as "Safety Clutter"—is just as critical as creating new ones. Ultimately, safety is about stopping the practice of counting failure and starting to measure capacity.
📚 Resources & Contact
To learn more about Dr. Erkal’s research on predictive safety, check out the following resources:
* Website: CSRA*
* LinkedIn: Elif Erkal, PhD
*All of their research and resources are publicly available and free to access, including their published papers. Since the work is publicly funded, listeners are able to explore the papers, literature, and additional resources available on their website.