A predictive shift for mine maintenance

RIO Analytics and OneWatt, innovative companies showcased in the Prospect Mining Studio cohort, are taking unique approaches to a fast-growing mining technology, shifting from the old method of preventative maintenance to a future of artificial intelligence-enabled predictive maintenance.

Arnhem, Netherlands-based OneWatt is utilizing AI for predictive maintenance in a remarkably interesting way, through sound.

Predicting and detecting faults in industrial equipment, OneWatt literally listens to motors using predictive AI with the company's embedded acoustic recognition sensor (EARS).

This non-intrusive solution saves industrial users, plant managers, and owners from unplanned downtime, revenue losses, and unproductive maintenance.

Out of Rio de Janeiro, Brazil, RIO Analytics is also seeking a way to use AI for predictive maintenance and has created a platform for just that.

By combining advanced industrial analytics and AI, RIO analytics has built a digital platform capable of predicting failures, reducing downtime, and increasing operational efficiency.

With perhaps the best approach to introduce cutting-edge technology into the mining industry, what better way than to keep existing technology and equipment functioning at peak condition for optimum performance.

By understanding the underlying "health" of a hard-working earth moving machine, a mine operator can determine if a corrective action is required before detrimental failure or damage occurs.

Thus, saving time, money, and potentially lives.

As more and more companies begin to advance into this sprawling avenue of unexplored possibilities with AI, the technology can only become more effective and RIO and OneWatt are paving that road.

OneWatt's EARS

Predictive maintenance relies on specific sensors to measure various machine parameters.

These sensors often track the performance of crucial components to show projected issues based on aggregated data.

This, in turn, allows technicians to more accurately anticipate when repairs will be needed based on real-time data from the machines themselves.

Presently, sensors that monitor for vibrations and temperature are used to track real-time issues – machine vibration are often an indication of imbalanced, misaligned or worn parts; and overheating or even lack of heat can indicate some kind of degradation or malfunction.

As electronic systems have gotten better at alerting technicians of possible repair, ultrasound sensors are even being used "see" into areas otherwise inaccessible.

OneWatt's EARS have taken the vibration method in a different direction. Using a custom non-invasive sensor, paired with a machine learning algorithm, the company's embedded acoustic recognition sensors can use frequency analysis to detect irregularities in an engine without any complex additions to the existing equipment.

With resonant frequency, OneWatt can determine equipment health via "listening" to its sound.

With such a truly imaginative and innovative method to approaching predictive maintenance, its exciting to think what other technologies may be developed in the future to tackle this ongoing and ever-present issue.

RIO Analytics

Besides preventing equipment failure, predictive maintenance can also lead to more efficient maintenance and eliminate unnecessary inspections. This strategy makes it easier to schedule repairs and order parts before a problem arises

The ability to schedule maintenance for the most convenient time, before an emergency, also allows for less planned and unplanned downtime.

RIO Analytics has created a fully integrated system – industrial equipment being monitored, AI, and a digital platform for navigating the metadata of monitored systems – to provide a complete approach to predictive maintenance.

The equipment branch focuses on the dynamics of existing systems as well as structural integrity and the work the equipment is undertaking. This includes the onboard power systems, the electro-mechanical behavior, as well as the monitoring of vibration mechanics, completing the full battery of inspection data to prepare for any problems.

RIO's AI algorithms are set up and trained to predict failures and increase efficiency, each modeled differently for individual assets.

RIO's AI include statistics and correlations, clustering of similarities, detection of outliers in present data, classifications and pattern recognitions with finally non-linear regression to further guess outcomes not assumed under any of the former methods of AI.

All of this is also supported with the continuous upgrade of the already existing algorithms.

Finally, RIO's KAIROS digital platform merges state of the art AI algorithms with existing industrial equipment knowledge and delivers data to the client.

KAIROS is equipped to streamline the entire predictive maintenance process all with the AI running in real-time and capable of pushing notifications through a dashboard that tracks data analytics, predictions, predictive warnings, and alarms.

Artificial intelligence has opened many new doors for the mining sector, with the work being undertaken by advanced computer systems freeing up time for further endeavors.

With RIO Analytics and OneWatt joining the ranks of companies introducing AI technologies to the mining sector, it is only a matter of time before new becomes the norm.