Infosys addressed the requirements of the mining company with an advanced vehicle telematics solution. We combined the Infosys Information Platform with Hadoop storage, Spark / Hive functions, R analytics engine, web visualization tools, and statistical modeling in the solution. The extendable architecture facilitated global deployment across assets from diverse mining and industrial equipment manufacturers, including Caterpillar, Komatsu, Atlas Copco, and MacLean Engineering.
Infosys partnered with Cisco to install a wireless sensor network at underground mining sites for uninterrupted streaming of asset data and monitoring of critical parameters. We mounted a telemetry device on each vehicle to log and transfer live data from the Wi-Fi-enabled underground infrastructure to the remote command center. The Infosys Information Platform ingests fleet telemetry data and shares it with the event processing and analytical engines.
The event analysis engine processes inbound data and monitors fleet performance indicators such as cycle times, speed, acceleration, and braking. It identifies Load-Haul-Dump (LHD) events and presents trend views / analytical reports that help fleet managers better manage operations. The Infosys event engine processes data three times faster than traditional models for event calculations.
Our solution applies big data analytics to mine business insights for operations, materials movement and safety incidents. It evaluates patterns in fleet telemetry and sensor data to correlate diverse operational issues – fuel consumption and maintenance events, driver behavior and fuel consumption, fleet maintenance and accident rate, etc.
Infosys developed an intuitive dashboard to manage and report mining operations, vehicle utilization, fleet performance, and safety incidents.
The menu enables users to select the date and shift, parameter and machine to compare and analyze performance across timelines, variables and LHD events. Asset / fleet summary reports can be exported in different file formats including Comma-Separated Values (CSV) and Microsoft Excel, and printed. Authorized users can configure application features such as thresholds for LHD events and color code for event detection.
Our ‘incident playback’ trends chart plots time series data based on user inputs, with minimal latency. It supports normalized trend analysis for event detection across 82 variables. User-controlled plotting features include dual axis, zooming, windowing, panning, and rolling