A look at Safety Analytics

Over the past few decades, safety performance at mines has greatly improved and rates of injury have decreased. At Goldcorp Inc., one of world’s largest gold mining companies, injury frequency rates have dropped 75 per cent over last 6 years, while over the same period, fatality rates remain basically unchanged. Adding rules, procedures and increasing compliance requirements have reduced injury rates, and while fatality rates remain very low, more can and should be done.

Goldcorp realized it needed an innovative approach to address this critically important issue that has faced mines for decades, and, in particular, understand why their fatality rates have remained stubbornly flat. They were eager to dig deep and understand all the various types of precursors that have and could potentially contribute to accidents and injuries. With more information about incidents, Goldcorp knew it could do more to prevent them.

The mining industry as a whole is working towards making mining as safe as possible. With this industry-wide priority in mind, Goldcorp began working with Deloitte’s Advanced Analytics team. Organizations have historically analyzed internal and external data to provide insights into key business challenges, such as improving operational efficiency and productivity, strengthening customer relationships, using asset more efficiently, streamlining production, and solving complex logistical problems.

Today, analytics are increasingly being applied to safety issues in order to generate actionable insights through the identification of behavioural or operational patterns that potentially contribute to safety risks. The Deloitte team has been working with mining companies to apply this analytic insight. Using best practices garnered from these projects, Deloitte deployed a unique smart safety analytics methodology — the first such program in Canada — to analyze Goldcorp’s data.

For this project, Deloitte focused on a single large Goldcorp mine, employing more than 1,000 people. Deloitte looked at a variety of Goldcorp’s existing data and corresponding external data. Over the course of three months, Deloitte built an analytic solution that looked back at 2,000 incidents over five years and tracked more than 1.8 million employee days worked. The data went beyond standard incident data to include operations data, weather, production, demographics, economic factors and more, totaling more than 792 million data points.

“We were trying to understand what we didn’t know. Without the analytics we were just making assumptions, but the analytics showed us the facts,” said Paul Farrow, Senior Vice President, People and Safety, Goldcorp. “By putting all this data together and applying visualization techniques, patterns began to emerge. We noticed trends that hadn’t been on our radar before. We were seeing causal factors that were contributing to incidents at the mine for the first time.”

As the smart safety analytics methodology was applied, one pattern that emerged was a relationship between compensation and injuries. Other interesting trends also came through quite strongly in the data, including relationships between injury rates and age, tenure and job roles. With this knowledge in hand, Goldcorp is now exploring procedures and management training that will help to address these issues.

Marital status also tended to make a difference in terms of number of incidents affecting a particular group of miners, with single men exhibiting higher incident risk. While there could be many reasons for this correlation, the insight provided by the analysis was empirical confirmation that supervisors and managers need to know their crew members and who might be more likely to take unnecessary risks.

“The findings of this project are driving behaviour change at the site,” said Farrow. “Now that we know that certain incidents have particular characteristics, we can take action in advance and before an injury or a fatality occurs.”

Technology advancements have provided organizations the opportunity to digest immense amounts of data from a variety of sources, it’s essential that this be transformed to information which can then be translated into intelligent actions. The true value proposition of safety analytics will be the understanding of what signals the data is showing and executing necessary change based on those insights. Taking operational action leveraging analytics is innovative and can accelerate the transformation safety in the mining industry.

“Based on this initial project at one mine, we are looking to expand this program company-wide,” added Farrow.  “In mining, safety really happens on the operations side and this allows for the deep analysis that simply wasn’t possible before.”