One of our priority goals for 2019 was to “use innovation to increase the effectiveness of our safety hazard assessments.” Our Structured Resource Allocation project used machine learning and sample plans to more effectively target our assessment resources on high hazards.
Machine learning allows systems to automatically learn and improve from experience, such as prioritizing assessments to identify high hazards. Sample plans build on these efforts by helping to measure the current status of high hazards across a specific population.
In 2018 we developed a pilot project that used machine learning to help prioritize gas and electrical installation permits. Once fully implemented, the results showed an improved ability to predict high hazard electrical and gas installation work. Due to its success, we implemented this model to gas and electrical operating permit assessments.
There were eight sample plans conducted in 2019 that contributed to:
- supporting our Compliance and Enforcement program;
- validating the machine learning model; and
- investigating various hypotheses related to specific industry areas.
Conducting these sample plans provided insight into how we improve the way we measure certain safety hazards and the impact of machine learning in the gas and electrical technologies.
Those who own, manage, install, maintain and operate regulated technical equipment are ultimately responsible for the oversight of their equipment.
The Structured Resource Allocation project is an example of how Technical Safety BC continues to improve its assessment model to better serve clients.