Intelligent minerals processing powered by AI and IoT
1 like3,174 views
This document discusses how to use AI and IoT technologies to optimize comminution circuit uptime and performance. There are three key levers for process improvement: 1) stabilize processes to reduce variability, 2) optimize processes to increase production rates, and 3) maximize asset availability and uptime. A case study describes a pilot project using cloud-based predictive maintenance on cone crushers in an African mine. Data from sensors was analyzed to detect anomalies and potential failures in order to prevent downtime events and unplanned shutdowns. The benefits of these technologies include eliminating unplanned downtime, optimizing circuit operations, and shorter/safer planned shutdowns.
Intelligent minerals processing powered by AI and IoT
- 1. © Metso Jani Puroranta Chief Digital Officer Metso Corporation Mines and Technology Europe May 30, 2018 How to use AI and IoT to optimize comminution circuit uptime and performance
- 2. © Metso Three key levers to process improvement in minerals processing and potential new digital solutions Pilot case study: Cloud-based IoT solution for predictive maintenance of cone crushers in an African mine 3 How to approach improving crushing circuit availability and performance with cloud-based technologies and AI In this presentation you will hear about... 2
- 3. © Metso Sources of lost production in minerals processing Productiontonnage manual operation planned shutdown process drift emergency shutdown upset conditions good control Actual tonnage Lost production potential Time set point 3
- 4. © Metso Three key levers to process improvement 4 1. Stabilize Reduce process variability 2. Optimize Increase production rate 3. Maximize Asset availability & uptime
- 5. © Metso How to approach predictive maintenance of the crushing circuit 1. Understand the problem 2. Identify root causes 3. Detect operational anomalies 5
- 6. © Metso Mean-time-to-repair(MTTR) Incident frequency Cracked mainframe Burned bushings Primary gyratory Screens Other Damaged shaft Cone crushers Low lube pressure High lube temp Damaged head Spider bushing damage Conveyor belt damaged Damaged chute Missing/broken mesh Broken mechanism What are the actual root causes? More data is needed. Jack-knife analysis 1. Understand the problem How to approach predictive maintenance of the crushing circuit 6
- 7. © Metso Fishbone diagram 2. Identify root causes Environment Material People Machine Method Measure Burned bushings Ignoring machine alarms Wet ore Packing condition Maximize shift tonnage No good measurement to indicate special process condition Changing ore properties How to approach predictive maintenance of the crushing circuit Could we develop an algorithm to detect and even predict the development of packing condition in real-time? 7
- 8. © Metso P-F curve concept 3. Detect operational anomalies Failure Agent Early stage Potential Failure Anomaly Agent P Functional failure P Late stage Potential Failure F Monitors already known patterns and early indicators of failure Identifies a steady-state operating pattern and looks for deviation F A How to approach predictive maintenance of the crushing circuit 8
- 9. © Metso Pilot case study Metso Nordberg® MP cone crushers in African mine 9 Three cone crushers connected to the cloud
- 10. © Metso Pilot case study Metso Nordberg® MP cone crushers in African mine Understanding crusher data Co-operation is key Data engineers Mechanical engineers Process engineers Reliability engineers Control engineers 10
- 11. © Metso Failure Detected a leakage in the hydraulic system a week in advance Trained to detect early damage to crusher head based on an actual failure Anomaly Agent Failure Agent Two types of predictive agents deployed FA Pilot case study Metso Nordberg® MP cone crushers in African mine 11
- 12. © Metso CIM 2018 | Jani Puroranta | Metso Analytics Process engineering Condition monitoring Predictive maintenance OEM support Root cause analysis Analytics, big data and AI in crusher optimization 12
- 13. © Metso Tons of sellable product Current tonnage Benefits Cloud-based technologies and AI in comminution 13 Elimination of unplanned downtime prevented downtime events less severe downtime events planned (vs unplanned) interventions Optimization of circuit and operating parameters % sellable product % target size recovery rate Shorter, safer and less frequent planned shutdowns more regular hours (i.e. less overtime) availability of right people availability of right parts
- 14. © Metso A sneak preview into the future 14
- 15. © Metso Maximizing equipment availability and uptime with cloud-based technologies and AI. Combining big data analytics and engineering expertise to understand complex crusher data. Co-operation with the OEM to improve availability and uptime means faster results. Summary 15
- 16. company/metso metsogroup metsoworldmetsoworld metsogroup www.metso.com