Abstract

Mobile mining equipment has been one of the most demanding application areas for AI-driven analytics solutions. The motivations for mining operations are clear: “Asset Health” analytics for this type of production equipment has the potential to significantly reduce downtime and maintenance costs, while increasing availability and overall operational effectiveness. However, delivering value to the end-user from such analytics solutions has proven to be challenging.

This paper presents recent case studies from successful deployments of Rithmik’s Asset Health Analyzer (AHA) solution suite, including examples of diagnostic maintenance, predictive maintenance, and fuel efficiency. These case studies demonstrate that the key enablers to success extend far beyond the capabilities of the AI and analytics engine.

The first prerequisite for delivering value is understanding what data sources are available, coupled with the ability to evaluate (and if necessary, remedy) data quality – including for “off machine” data sources such as fleet management systems and work order systems. This requires complementing data science expertise with very specific domain expertise related to the operational mining context. The next key enabler for success is understanding the organizational structures and workflows at the specific mine site, to clearly define what level of “actionable insights” (e.g. alarms, trends, forecasts) are to be generated, which key personnel need to receive those insights, in what form, and when. This requires familiarity with mining operations and the ability to integrate into the information systems preferred by end-users.

Successful, sustainable, deployments of solutions such as Rithmik’s AHA hinge upon effective change management, and deployment risk is significantly mitigated if the mine site has in place, or adopts, a continuous improvement methodology for its overall reliability and maintenance activities (e.g. RCM). Another significant benefit from adoption of such a continuous improvement methodology is that it enables and embeds the retention of historical knowledge of failure modes and appropriate responses, through formalized documentation processes. This knowledge retention is of increasing value, given the demographic challenges of the mining workforce and the increasing rates of turnover of experienced personnel.