Client story Leading with Industrial Internet-of-Things to Drive Data Science Initiatives

Facts at a glance

Client industry:

Mining

Challenge:

The client looked to drive mining operational efficiencies to improve throughput and minimise lost opportunity costs, through data science initiatives.

Solution:

• Data Science Platform based on Azure
• Connecting 100,000 machinery sensors to Azure IoT into the Data Lake

Results:

• Enabling client’s data scientists to leverage the platform to uncover machine learning algorithms to drive preventive maintenance outcomes

Solution area:

Insight’s Digital Innovation solutions help clients digitally transform their business operations to optimise operations.

The concept and the challenge

How informed decision making is saving millions of dollars

Newcrest Mining embarked on a digital transformation program which turned to Industrial IoT and Advanced Analytics to enable early detection of constraint conditions that prevents its mining sites from achieving peak production rates.

The key objective was to build a platform capable of capturing and processing 40 billion records of machine sensory data, from 100,000 sensors across their 24/7 mining operational internationally, captured at 2-3 second intervals. Only a robust and scalable platform could handle this large amount of data, and allow for Newcrest’s Data Science partners to explore and identify predictive maintenance algorithms that would avoid production downtime – saving millions of dollars.

The solution

An internet-of-things platform solution

Insight architected and implemented Newcrest’s Data Science Platform, built on Microsoft technologies, to transfer the massive amount of sensor data into a data ‘lake’, with best-in-class in-memory processing capability that could allow for quick data preparation and discovery, and to leverage machine learning models from the R & Python communities.

The use of this platform, among other initiatives such as crowdsourcing solutions, is why Newcrest is regarded as a leader in innovation for the mining industry.  The company dedicates significant resources to scouring and assessing unconventional approaches to complicated site challenges.

"The platform provided a rich set of data services with the flexibility to handle the many transport, processing and visualisation scenarios Newcrest requires."

“Ignia [an Insight Company] has been a strong partner, helping us set the foundations for the future by looking into newer-technology edge based computing, advanced pattern recognition and integration to cognitive computing. This is helping us to achieve further production optimisation and reduce unplanned outages of our core processing facilities.” – Travis Ray, Enterprise Architect, Newcrest

Another great outcome

Robust solution, designed to last for the future

Newcrest has achieved extraordinary operational efficiencies through data science initiatives such as the Conveyor Belt plant maintenance prediction model.

With the new ability to obtain such a large dataset, data science tools have developed a predictive maintenance model as one example of how the platform has become a key enabler for millions of dollars of future recurring benefits.

"By using Microsoft HD Insight clusters, 4 years’ data from the Cadia East underground conveyor belts’ 90 tags were summarised to 15 second intervals and extracted."

Newcrest’s Chief Information & Digital Officer, Gavin Wood explains, “You’d struggle to find many other IT departments paying for themselves two times over, and we do. A lot of that is built on taking the data and using visualisation and dashboards to empower people to take decisions to drive the business harder. We delivered over US$50 million worth of future recurring benefits in the last financial year from data science initiatives and process control automation, optimising our operations by increasing throughput and recoveries. We are targeting continuation of such benefits in this financial year.”

Maximise your ability to achieve operational efficiencies. Consider a modern data platform to drive predictive maintenance.