Predictive Asset (Device/Equipment) Maintenance

Business Problem

Across industries unexpected and unplanned downtimes disrupt factory schedules leading to lower productivity and risk to on-time deliverables. Organizations constantly explore opportunities that allow scheduling of corrective maintenance to prevent any unexpected equipment failures.

Solution Description

System/application based on predictive analytics that helps reduce unscheduled asset downtime with the following features.

  • Real time alerts for equipment failure
  • Indications for type and possible time of failure
  • Email/messaging to respective personnelto take corrective action
  • Standard BI reports / visualization for equipment performance / failures

Solution Workflow

  • Build predictive engine for equipment / device failures basis historical information
  • Collect equipment’s monitoring information at regular time intervals and feed predictive engine
  • Predictive engine to provide alerts on any possible equipment failure and insights on type and time of possible failure
  • Communicate alert information to respective personnel for corrective action

Leveraging HANA capabilities

  • Content Creation/Modification/Deletion
  • Feeds, News and Alerts
  • Advanced search using Date, Author, Tags, and Keyword etc.
  • Discussion forum
  • Sharing, Messaging and Tagging
  • Customizable Hierarchy

Leveraging Open Source Capabilities

  • Application would be hosted on the cloud and will need to support multiple clients with multiple installations / plants and would leverage on HANAs

    • Multi Tenancy and cloud deployment
  • This is the typical relationship that exists – Component → Devices → Equipment → Multiple Equipment → Plant. Monitoring information needs to be collected at real timeat a component level if configured so. This would require real time integration of significant volume of records and would leverage on the following HANA components.

    • Data Services
    • ESP
  • And leverage on the following features

    • Column & Row store
    • Partitioning
    • In-memory compression
    • Active/Passing & Data Aging
    • Multi / Core parallelization
  • Predictive engine that runs robust algorithms to match against historical records to determine probability of failure at a component / device / equipment level in real time and would heavily leverage on following HANA components

    • Predictive Analytics Library
    • “R” HS Integration
    • Parallel Calculations
  • And leverage on the following features

    • Column Store
    • Analytics on historical data
    • Parallelization
    • In-memory
  • Standard BI reports / visualization for equipment performance / failures which leverages OLAP features of HANA
  • Communicate alert information to respective personnel for corrective action would leverage on the following HANA components

    • OData
    • SQL

Solution Benefits

  • Predict equipment/device failurethereby ensuringhigh asset availability and performance
  • Reduce operational costs and enhance asset productivity
  • Increased and timely yields from processes leading to optimization of quality and supply chain processes