www.wipro.com Application of Big Data Solution to Mining Analytics Sandipan Chakraborti Senior Architect ENU

www.wipro.com Application of Big Data Solution to Mining Analytics Sandipan Chakraborti Senior Architect ENU

Table of Content 03... Introduction 03... Big Data & Analytics Across Mining Functions 05... Interventions Across Mining Processes 06... Where Lies the Value 07... About the Author 07... About Wipro Ltd.

Introduction Big Data Analytics is now a big blip on the radar of the Mining industry. In a recent survey that included 10 of the Top 20 global mining companies, the Mining Journal said that Big Data Analytics would spur the next wave of efficiency gains in ore extraction, analysis, transportation and processing by enabling faster and better informed decisions at all levels. In a competitive market, every effort to improve margins using operational intelligence is necessary. That is why analytics is expected to play a major role in driving better asset utilization, boost productivity, and address material flow delays. Helping achieve this goal are sensors embedded across mining operations. These sensors are generating vast amounts of geoscientific, asset condition and operational data in real time. Improvements in Wi-Fi and 3G/ 4G-LTE speeds are enabling real-time collection of data from the extraction point right up to the final transportation of ore to plants. This data can be analyzed using massively parallel processing and faster distribution of intelligence to stakeholders. It is possible to do this because modern Big Data platforms can assimilate vast amounts of heterogeneous, real-time inputs from multiple sources. These, in turn, extract real-time predictive and prescriptive analytics to drive operational excellence. Big Data & Analytics Across Mining Functions Data sources in the Mining industry may be classified as either direct or indirect (ancillary) measurements. Direct measurement sources are those taken by instruments such as conventional geodetic surveys and GPS. Indirect sources refer to systems that collect data as a by-product of processes or operations such as Fleet Management Systems, SCADA or DCS data, blast hole drills and geo modeling data. To improve ore recovery, an ore body modeling technique is used. The model provides geological patterns that determine drill holes. The key to taking the right mining decisions is, therefore, the availability of accurate data from multiple systems combined with real-time (or near real-time) analytics (see Figure 1). These decisions can be applied to mining exploration, production and operations. They can also be used to monitor and report metrics and KPIs. Additionally, they serve to identify root causes for operational bottlenecks such as unscheduled truck maintenance delays, long queuing time of trucks and LHDs, delays in lab samples undergoing quality control and batch processing etc. 3

Consumption Layer Adhoc Discovery and Visualization Business Process Management Mining Remote Operations Control Room Real-time Monitoring Reporting Engine Transaction Interceptor Real-time Data Self Service / Custom Business Alerts Operational KPIs Navigation Query Reporting Dashboards Analysis Layer Complex Event Processing Machine Learning / NLP Real-time Scored Results Decision Management Recommendation Engine Model Management Predictive Model / Statistical Model Big Data Messaging & Storage Layer Data Acquisition Data Digest Distributed File Storage Big Data Source Structured, Semi-structured and Unstructured Text Images Audio Video Spatial Temporal Document Relational Data Domain Entities Direct Data Store Mining OT Systems (Ancillary Systems) Data Stores ECM ERP GPS Data Geoscientific Data Dispatch Drill & Blasting FMS Rail Track System SCADA Asset Condition Monitoring Plant Control System Ore Geomodeling Real-time DB Data Warehouse Excel Word Digital Assets Production Planning Supply Chain Management Smart Devices Health & Safety Apps Tripsheet Management Figure 1: Big Data Analytics Solution Framework Aside from providing insights for decision-making, the Big Data Analytics Platform can also provide prescriptive solutions around decisions (for an example see Figure 2). Storage Impact of LHD Cycle Time on Throughput Rate Processing & Consumption Usage Mining Material Flow Predictive, Prescriptive Data Analysis Correlating the causal variable patterns such as loading time, hauling time, dumping time, idle time and queuing time on the overall Average Cycle Time over a period New Existing Data Source Real Time Timing Batch which would impact the Throughput Rate. Apply historical as well as real-time data analytics to determine patterns which can help in predicting production throughput. Figure 2: Big Data Analytics on Mine Material Flow 4

Interventions Across Mining Processes Material process flow plays a big role in the mining value chain. This includes analyzing impact of unscheduled events owing to mechanical breakdowns of LHDs, trucks and critical transportation medium, queuing time, and such overheads. There are a number of other causal variables that can be analyzed for impact on production throughput on a daily/monthly basis using techniques such as Machine Learning, Continuous Pattern Matching and Statistical Predictive Model. Big Data Analytics Platform, equipped with these models, can leverage the value, volume, velocity and variability of data, delivering several benefits across extraction, intermediate transportation and final transport to plants. Figure 3 shows the causal data used at each process step to improve operational effectiveness and enable higher ore yields. Causal Processes Extraction Intermediate Transportation Final Transport at Plant Benefits Optimized process flow Casual Data Casual Data Casual Data On-time delivery of minerals to plant LHD Data Productive Data Extraction Points Shift Cycle Time Operator Ore Pass Operators Piques Data Cycle Time LHD Data Operator Ore Pass Dumping Points LHD Data Equipment Condition Higher throughput rate and cycle time Explicit view of performance alignment with Production KPI Value Improved Production Casual Analysis to identify process bottlenecks Control Variability in the process Variability Heterogeneity of data from Dispatch Control Systems, Fleet Management Systems and SCADA Velocity Frequency of data change is real-time ETL batch process driven Volume Historical data amounting to several terabytes for analysis Figure 3: Causal and Correlation Analysis using Big Data Some use cases where Big Data Analytics Platform can come in handy are: LHD performing loading hauling dumping cycles from Extraction Point A to Dumping Point B for a given mine site in the transportation cycle time Mine Manager needs to identify and correlate Typical causal variables impacting daily / monthly production variations (a) Effective Operating Hours of the LHD (b) Available Hours (c) Scheduled Hours (d) Utilization % Extraction Point Dumping Point LHD Figure 4: Mining Use Case A 5

Figure 5: Mining Use Case B LHD performing loading hauling dumping cycles from Extraction for a given mine site in the transportation cycle Mine Manager needs to identify and correlate the following Haul Truck Performance Variables behind the Daily / Monthly Production variances (a) Tons Moved (d) Average Load Cycle (b) Throughput Rate (e) Total Cycle Time (c) Operating Hours (f) Average Cycle Time Cost of unplanned and mechanical outages of Trucks is a big expense on mining companies. Typical parameters that drive the maintenance cost are as follows to name a few of the variables (a) Engine Oil Pressure (b) Engine Oil Temperature (c) Hydraulics Oil Temperature (d) Transmission Oil Pressure (e) Transmission Oil Temperature (f) Coolant Temperature (g) Break Changing Pressure Using Big Data Analytics platform it is possible to identify machine patterns and predictive models to do proactive maintenance of Trucks. Figure 6: Mining Use Case C The mobile drill rigs of the future resembles a mobile surveying and sampling laboratory which can collect, analyze and access huge volumes of complex geochemical and geophysical data. The data can be synced with the central server for validation. QA/ QC routines built into data collection mechanism ensure data quality problems are identified at the source. The adoption of Big Data platform ensures complex and near real-time geochemical/ geophysical data is able to be processed and analyzed. The interpreted results from analysis is communicated near real-time to survey geologists. Decisions Communicated to Gvveologists QA/QC Geochemical Data Assay Results Analysis & Intereption Near Real-time Dashboard Mobile or Drill Rig Exploratory Drilling Data Digital Instruments Sync Complex and Voluminous Datasets Big Data Platform Central Server Drill Rigs of the future Where Lies the Value Figure 7: Mining Use Case D The Mining industry can derive several critical business benefits from Big Data Analytics. These include: Ensuring continuous flow of material from ore extraction point to the processing plant Maximizing ores hauled by optimizing bottlenecks in production Reducing non-productive time between unit operations such as unscheduled maintenance, delays, wastage and waiting time Helping management make informed decisions on the as-is production process, covering the value chain from extraction to delivery at plants and beyond Providing on-the-fly assay results and interpretation analysis to field geoscientists to take informed decisions For organizations considering such a platform, ensuring a low Total Cost of Ownership without vendor lock-in, with the ability to scale horizontally should be major considerations. 6

About the Author Sandipan Chakraborti is working as a Senior Architect in Wipro s Energy, Natural Resources & Utilities business. His work involves architecting presales solutions, realizing architectures through projects and engaging with customers for proof of concepts. Sandipan has over 16 years of experience in areas of consulting, presales and delivery. Prior to working with Wipro, he was a Principal Consultant in the strategy and architecture unit of a leading global consulting organization. He can be reached at sandipan.chakraborti@wipro.com. About Wipro Ltd. Wipro Ltd. (NYSE:WIT) is a leading information technology, consulting and business process services company that delivers solutions to enable its clients do business better. Wipro delivers winning business outcomes through its deep industry experience and a 360 degree view of "Business through Technology. By combining digital strategy, customer centric design, advanced analytics and product engineering approach, Wipro helps its clients create successful and adaptive businesses. A company recognized globally for its comprehensive portfolio of services, strong commitment to sustainability and good corporate citizenship, Wipro has a dedicated workforce of over 160,000, serving clients in 175+ cities across 6 continents. For more information, please visit www.wipro.com; info@wipro.com 67

DO BUSINESS BETTER CONSULTING SYSTEM INTEGRATION BUSINESS PROCESS SERVICES WIPRO LIMITED, DODDAKANNELLI, SARJAPUR ROAD, BANGALORE - 560 035, INDIA. TEL : +91 (80) 2844 0011, FAX : +91 (80) 2844 0256, Email: info@wipro.com North America Canada Brazil Mexico Argentina United Kingdom Germany France Switzerland Nordic Region Poland Austria Benelux Portugal Romania Africa Middle East India China Japan Philippines Singapore Malaysia South Korea Australia New Zealand WIPRO LTD 2015 No part of this booklet may be reproduced in any form by any electronic or mechanical means (including photocopying, recording and printing) without permission in writing from the publisher, except for reading and browsing via the world wide web. Users are not permitted to mount this booklet on any network server. IND/BRD/DEC 2015 FEB 2017