Non-profit Safety Regulator uses Machine Learning to Improve Public Safety

Print Friendly, PDF & Email

Machine Learning Case Study

Innovation in data science and artificial intelligence / machine learning are more often than not associated with the latest news out of Silicon Valley. However, there are a growing (and arguably more important) number of applications of these incredible technological innovations in the non-profit sector, which are creating direct benefits for everyday people in communities across the globe.

One example is Technical Safety BC, an independent, self-funded organization that oversees the safe installation and operation of technical systems and equipment across the province of British Columbia in Canada. The not-for-profit organization recently partnered with data science software maker Dataiku to introduce more sophisticated machine autonomy to their hazard identification process. The partnership enables Technical Safety BC to build machine learning and advanced analytics-based solutions faster and more accurately, allowing the company to better target areas of high risk.

In safety regulation, conducting physical assessments is costly and false positive inspections (i.e., sending safety officers to inspect sites where no apparent risk to the public or asset owner exists) can result in significant opportunity costs each year, where those same resources could be better utilized within the safety system. Therefore, finding a way to more accurately predict hazards is of high strategic value to organizations like Technical Safety BC, and creates greater safety benefit to the public.

By introducing more sophisticated machine autonomy in the risk assessment process, Technical Safety BC aimed to find more high hazard sites while operating at the same resourcing level. Some of the challenges they faced included: uncoordinated heterogeneous data sources; data quality; speed of collaboration; and training challenges in the use of machine-recommended predictions.

To address these challenges, Technical Safety BC created a dedicated data analytics and decision science team responsible for integrating advanced analytics into all parts of the organization. The team chose Dataiku as the tool they’d use to bring efficiency gains to their data processes. Using Dataiku, Technical Safety BC is able to quickly prototype, test, iterate, and deploy innovative, data-driven solutions.

Dataiku’s collaborative environment enables the team at Technical Safety BC to easily leverage work done by colleagues. For instance, teammates can reuse models repeatedly and easily share queries that each individual team member might otherwise have had to write and rewrite for similar projects. These efficiency gains free up the team to spend more time devising innovative new ways to experiment with various models and provide value to their clients and to the public.

For example, Technical Safety BC recently used Dataiku on a project that leverages machine learning to identify common features and signals relating to risk factors. Built on the latest machine learning technologies, the new models developed in this project adapt quickly to reflect any emerging risks and to help Technical Safety BC shift resource allocations accordingly based on the most current knowledge. With the ability to rapidly prototype using Dataiku, the team used A/B testing to check the performance of the new machine learning models in predicting high hazard sites.

Since deployment of these new machine learning models developed with Dataiku, Technical Safety BC’s predictive performance improved by 85% for models predicting hazards with electrical technology compared to previous methods used. Machine learning with Dataiku also significantly reduced the total number of mandatory inspections for Technical Safety BC, which will help optimize safety officer time so they can apply their time and expertise to potentially higher hazard situations and make a greater impact on safety. Moreover, the company has now developed a data science roadmap around the use of Dataiku, which is comprised of user engagement and empowerment, rapid prototyping, and testing of data products as well as an ethics roadmap to guide the use of such data products to deliver best-in-class safety oversight.

 

Sign up for the free insideBIGDATA newsletter.

Speak Your Mind

*