The AuOps Center will process these vast quantities of data from sensors, equipment and digital tools located across Barrick’s mines in Nevada to convert it into timely and relevant information. This will help all levels of operations, from operators to site general managers, to make data-driven decisions in real or near real time.

The centre will be the first of its kind and feature in-house analysts to interpret data and develop insights for site personnel to make better, faster and safer decisions.

Rob Neitzel, digital project owner for the AuOps Center, explained: “Analysts located at the centre will improve our decision-making by interpreting in real time how our actions will impact health and safety, the environment, production and costs, and communicate that to our people at site.”

The AuOps Center is not a traditional control centre that’s equipped with remote technology or Wi-Fi.

The centre will retrieve and analyse data then contextualise that data into a visual representation with an explanation for the analysts to produce actionable insights. This will help everyone understand in real time how the mines are performing against key performance indicators. This will also help them identify potential opportunities and risks before making decisions.

The centre will capture data on core functions from the Nevada mines such as mining, processing, maintenance and resource load. The data will be accessible through the Consolidated Data Platform, which acts as a cloud, and other data sources. Using advanced analytics and visualisation tools, analysts will then interpret easy-to-read visuals and graphics to understand how the mines are performing, the factors behind those results, and the impact they will have on various metrics, including cost and production.

Analysts can then decide who to contact at site with the relevant information and advise them accordingly. The operations team will also have access to the centre’s performance dashboards through their tablets, and will be able to communicate directly with the analysts for round-the-clock support.

Predictive analysis will flag potential issues such as variances in measured consumables or equipment nearing failure. For example, an analyst will receive an alert if a water pump has an 87% chance of failing in the next three weeks. The screen will display the estimated cost of fixing it now versus replacing it after it fails.

The cost savings could be substantial, even if budget is exceeded by a small amount. The analyst will then share this information with the relevant person, such as a site superintendent. If the superintendent chooses to fix the water pump now, her decision will be logged in the system. When she requests additional funds for her maintenance budget, her supervisors will know why.

Analysts will also identify, track and monitor health, safety and environmental concerns. So, if the centre detects a situation at site that could impact safety and health, an alert will appear on the analyst’s screen, dictating the severity and the next steps that should be taken using Barrick’s existing escalation checklist.

In addition, analysts will track performance across operating teams, identify the equipment used and monitor key performance indicators and performance metrics. This will help identify teams that are performing well and drive best-in-class operator performance.

For management, this means they will get an understanding of key performance indicators and what the predicted outcome would be each week, month and quarter if operations continue to execute in a consistent manner. The analysis will indicate potential opportunities and risks. For example, a report could show a two-day experimental blend at the mill resulted in higher production. But if that experimental blend is continued, the analysis will show a potential impact to the processing permit limit, which allows only a certain amount of gold to be produced each year. Management, however, could use this information to apply for a larger permit the following year.

Construction on the centre recently finished and the analytics section will commence implementation in August and gradually build on functionality until next spring.