Project software

BHP unveils major data science project – Software

BHP is applying data science to understand how it maintains machinery in its mines, hoping to save $ 79 million for this exercise alone.

The miner revealed late last year plans to establish a Brisbane-based Maintenance Center of Excellence (MCoE).

The MCoE will standardize maintenance systems and processes for BHP’s global operations, replacing the previous model of having 40 different maintenance organizations around the world, each with their own unique way of working.

One of the keys to the MCoE model is that it leverages data science techniques, such as machine learning, to understand how maintenance is performed at each site and where improvements can be made.

Like other projects since BHP relaunched its technology function earlier this year, the idea with MCoE is to create repeatable processes for its business operations around the world.

“We have gone to great lengths through internal and external benchmarking to understand in depth what has kept us from achieving our maintenance ambition in the past, so that we can ensure our success in the future,” said the vice president of maintenance, Brandon Craig.

BHP presented the first details of a five-year timetable for the application of its new maintenance methodologies to different types of fleets.

It focuses first on ‘higher value-added opportunities’, which for BHP have been improvements to the Caterpillar 793F haul trucks – a key part of its iron ore mines in Western Australia, but also used in its global operations.

Analyzes of engine performance from its worldwide 793F fleet have revealed “a great deal of variation” between mines in the life of engine components before they need to be replaced.

“The longer the life of the equipment and its components, the less physically we work on it and the less money we end up spending,” said Craig.

“By analyzing the causes of these differences, we can put in place solutions that elevate the underperforming to match the best, and in fact beat them.

“Our new data science techniques and the increasing automation of reliability engineering have largely solved what historically was a very human process and which has slowed us down considerably. This is a fundamentally new capability and a new set of tools for BHP.

“As a result, we can now do in one day what would have previously taken us weeks, and most importantly, it also creates new perspectives to understand all of our sources of loss. “

BHP said it has created machine learning algorithms to analyze component failure history and analyze engine component wear in real time, allowing it to “better predict failures” and schedule more maintenance. in advance, more precisely.

Craig has revealed the first results of maintenance improvements for the 793F at his Yandi mine in the Pilbara.

“We have analyzed all the data we have in detail and, based on this information, we have created new maintenance strategies to optimize safety, costs and performance, but not just for trucks – we have worked on all end-to-end maintenance. and the supply chain system that supports these trucks, ”he said.

The accuracy of truck maintenance planning at Yandi has improved dramatically: “from only 10% of jobs planned more than two weeks in advance being 85% accurate”.

BHP has also been successful in reducing instances of parts downtime when trucks need maintenance.

“Supply chain accuracy has improved from 13% of missing parts when we have to work or service these trucks, to just 1% of missing parts, and for scheduled jobs it is typically zero for hundred, ”Craig said.

The project had generated $ 5.5 million in savings for Yandi in fiscal 2017, and the availability of trucks for work at the mine was “well over 90 percent”.

In addition to optimizing the management of its global 793F fleet, BHP has also applied its MCoE capabilities to its Leibherr T282 trucks. The largest user of these is the company’s New South Wales ‘power coal’ operation in the Hunter Valley, which supplies coal to power producers.

During these operations Craig said the average time between fleet failures had improved “from 36 hours to 53 hours”.

The third type of fleet for which maintenance is optimized are the Caterpillar bulldozers.

In all three cases, BHP is expected to achieve savings of several million dollars in FY18.

The company is now starting optimization work for a range of excavators and processing plants around the world.

“We expect the performance improvement we have made to date will continue across the remaining fleet, equipment and processes,” said Craig.

Overall, the MCoE is expected to have a “significant impact on the safety, volume, costs and performance of maintenance capital” across all of BHP’s global operations.

“We plan to unlock more than 3.5 percentage points in [fleet] availability which we believe will translate into an additional eight million tonnes of iron ore, two million tonnes of coal and 45,000 tonnes of copper, ”said Craig.

“In addition, we aim to generate $ 700 million per year in recurring cost savings.”

These numbers are made possible by the scale of BHP’s maintenance operations: the company spends $ 3.5 billion annually on maintenance, “and that number is closer to $ 5 billion when you include capital. maintenance, ”said Craig.

“So overall, these direct expenses represent about a third of the group’s operating expenses,” he said.

“We have 10,000 people directly employed in maintenance across the group and when we factor in subcontractors, that’s over a third of our total workforce of 60,000.

“With more than 3000 fleet units and more than 1.7 million [maintenance] work done each year, we have a significant opportunity to leverage our scale through increased standardization of work to pursue much greater standardization of the fleet and to improve faster through greater duplication of tasks.

Source link

Leave a Reply

Your email address will not be published.