You are on page 1of 3

ARTICLE

Optimizing mining and metals maintenance


strategies with predictive analytics
Operational efficiency is a new industrial imperative
Worldwide, the mining and metals industry is using operational data to
improve asset operations and reliability. Many organizations have already
shifted from regular, scheduled maintenance to using real-time insights from
operations data to determine asset maintenance schedules. This condition-
based maintenance approach lowers costs and improves asset availability
and reliability.

While maintenance teams can view a real-time assessment of an asset’s


current health, the assessment alone provides no indication of future
events. Consequently, mining and metals companies are now deploying
predictive maintenance strategies. Predictive maintenance uses historical
and real-time asset data to train artificial intelligence and machine learning
models. Those models then project future outcomes, offering unprecedented
situational awareness, while preventing potential catastrophic issues and
optimizing asset availability.

Converting unplanned outages into planned equipment shutdowns and


improving asset reliability can increase operators’ margin by millions of
dollars. In addition, converting other unplanned downtime to preventative
maintenance further decreases costs by allowing the maintenance to be
scheduled during a lower margin window.
Performance Intelligence brings efficiency to Barrick Gold: Turning operations data into a
your data gold mine of insights
Traditional maintenance approaches use visualization Headquartered in Toronto, Canada, Barrick Gold is one
and alerting tools while predictive maintenance relies of the largest gold mining companies in the world. With
on advanced analytics. Making the transition can mines in 11 countries, Barrick creates a wealth of data
be difficult, particularly if the predictive models are across the enterprise.
developed in-house. Organizations may not have the
Extracting and refining gold in Barrick’s Cortez mine in
data science experience to build complex models, the
Nevada is a laborious, mechanically intensive process.
solution may not be easily scaled, and data may be
Machinery is expensive and so is downtime caused by
locked away in functional silos. In addition, models
any mechanical failure. To mitigate these costs, Cortez
require significant retraining and refinement before they
invested in a predictive maintenance strategy using the
can be operationalized.
AVEVA PI System. The mine deployed a mixture of IIoT,
Integrating an enterprise system of record with data analytics, sensors, and machine learning tools
mature predictive analytics software can enable to detect potential equipment problems before they
mining and metals companies to overcome many of reached the failure point.
these problems and deploy a predictive analytics
Barrick uses sensors and vibration monitoring software
project much more quickly. This combined approach
to analytics infrastructure gives mining and metals from Petasense IIoT sensors to gather information from
operations visibility across entire asset portfolios. It assets that are not inherently IoT-enabled. Machine
also enables users to leverage predictive analytics to learning algorithms use frequently collected sensor
proactively maximize operational efficiency, resiliency, data to generate an equipment health score and track
and sustainability–all of which are key outcomes for the declines that indicate a potential problem.
mining and metals industry over the next decade. Site engineers use PI System tools to integrate
Petasense data with other sources and to provide
Enable the digital twin and empower the clear, intuitive access to that data for a variety of
connected worker users. The first potential defect was identified within a
Centralizing data management means that mining month, and a single early fault detection for one piece
and metals companies can accelerate the time- of equipment saved the company $600,000. Overall,
to-insight for operations, maintenance, planners, predictive maintenance has enabled the mine to reduce
business analysts, and more. The single system of unplanned downtime and repair costs and improve the
record becomes a digital thread that connects all the overall efficiency process.
devices monitoring asset performance with a predictive
analytics engine. It is the foundation for many different Votorantim Cementos: Predictive insights
projects, from predictive and prescriptive maintenance optimize operations
of a single asset, tocreating a digital twin of an entire AVEVA Predictive Analytics provides early warning of
asset portfolio. Staff no longer have to track down equipment failure and abnormal operating conditions
performance data in disparate systems and manually that may go unnoticed with traditional maintenance
upload it to analytics software. They do not have to be practices. Votorantim Cementos, a Brazilian cement
software developers or data scientists to start gaining manufacturer, wanted to move from reactive to
insights from operational data. predictive maintenance strategies. The company
selected AVEVA Predictive Analytics because it created
Instead, anyone stakeholder can immediately analyze
a single source of reliable data and information and
operational data from across the enterprise. For
enabled machine learning-based insights. The solution
example, maintenance teams can start a condition-
was deployed on top of Votoranitim Cimentos’ existing
based maintenance project using mature analytics from
AVEVA PI System.
a single source of reliable operations data.

2 Optimizing mining and metals maintenance strategies


with predictive analytics
After a rapid deployment, the company quickly began Overall, the company reduced its maintenance-related
to see the benefits of predictive maintenance. Across costs while improving asset reliability and virtually
six sites, predictive-analysis driven “catches” avoided eliminating the need to make emergency purchases.
$23 million in corrective maintenance costs per site Using AVEVA Predictive Analytics, Votorantim
in the first year alone. Over time, AVEVA Predictive Cimentos cut its maintenance backlog by 10%,
Analytics enabled the team to not only manage boosted asset reliability by 6%, and enabled
immediate maintenance needs, but also plan and scale comparative analysis between plants to share and
processes, when needed, to minimize operational risk promote best practices.
and unscheduled downtime.

For more information about Predictive Analytics, please visit:


aveva.com/en/products/predictive-analytics

© 2021 AVEVA Group plc and its subsidiaries. All rights reserved.
AVEVA and the AVEVA logo are a trademark or registered trademark of AVEVA Group plc in the U.S. and other countries.
All product names mentioned are the trademarks of their respective holders.

aveva.com

You might also like