Explore solutions by selecting your industry

Industries

MaintenanceAI @GEMx

Artificial Intelligence for Maintenance in GEM

MaintenanceAI @GEMx

Artificial Intelligence for Maintenance in GEM

What it is

MaintenanceAI is an end-to-end reliability and maintenance system​ where workforce efficiency is optimized and failures are detected early to allow for seamless interventions; as a result, total asset availability is maximized while its total cost is minimized

Key domains are

Reliability & Predictive maintenance
  • Real-time, online asset monitoring
  • Early failure detection
  • Automated critical alerts / notifications
  • Supply chain integration with PdM platform
Digital work management
  • Workflow digitization / Mobile execution
  • Real-time performance management
  • Smart / automated scheduling
  • Real-time tracking of outages / turnarounds

How it works

Reliability & ​ Predictive maintenance
Digital work management

Description

  • Solutions that seek to reduce the occurrence and impact of equipment failures​ through early indication of anomalies, and optimize planned maintenance​ / turnaround activity
  • Solutions that digitize end-to-end maintenance workflows​ from initial work identification, planning & scheduling, execution, closeout and improvement

How these solutions work

  • Predictive models aim to identify complex anomalous behavior across sensor tags
  • Features driving anomalies are used to guide early intervention to avoid failures
  • Modelling asset health and criticality allows for planned work to be optimized
  • Solutions connect with master data from WO systems (SAP, Maximo etc.)
  • Modules for schedule optimization, ePTW and digital SOP’s etc. are provided
  • Solutions work across mobile devices to ensure single source of truth

Solution focus areas within these domains

  • Replacement CAPEX​ – optimization of capex spend decisions based on condition data and lifecycle costs
  • Prioritization of planned work​ – optimizing planned maintenance based on health & criticality
  • Imminent failure avoidance​ – detection of anomalies in sensor and notification data for early interventions
  • Remote monitoring centers​ – to enable deployment and adoption of predictive maintenance solutions
  • Work identification​ – ​ using in-field mobile devices for rich data capture allowing for data-driven prioritization
  • Planning and scheduling​ – planning of work packages, scheduling of activity and digital workflows for PTW
  • Execution and closeout​ – digital work instructions and SOP’s and automated work invoicing
  • Performance improvement​ – analytics applied to workflow activity data to drive continuous improvement

PdM and DWM are complementary within end-to-end maintenance processes

Maintenance phase
Role of digital work management solution
Role of reliability & predictive maintenance solution

1. Work identification

Frontline apps for operator rounds allow digital capture of observed equipment issues with rich metadata …

… Predictive maintenance health and anomaly detection models consume observation data logged by operators

2. Planning and scheduling

… Frontline maintenance schedules dynamically updated based on model prioritization

Health & criticality models used to prioritize planned maintenance tasks …

3. Execution and closeout

… Digital work instructions and interactive SOP’s are guided by PdM root cause analysis

Model feature importance and failure archetypes allow for automatic issues root cause analysis …

4. Performance scheduling

End-to-end digital workflows allow performance metadata to be captured and analyzed to identify improvement opportunities …

… Critically models consume job execution performance data to re-evaluate predictions on failure consequences

Value levers

  • Reducing unplanned downtime​ – anomaly detection for imminent failure mitigation and avoidance
  • Prioritization of planned maintenance​ – using health and critically models to optimize levels and allocation of PM work
  • Increasing wrench time – increased frontline efficiency for smaller and more agile frontline teams
  • Reducing re-work​ – digital work instructions ensure less human error and costly procedural mistakes
  • Providing actionable RCA – predictive models provide insights into health drivers for more targeted root cause analysis
  • End-to-end job delivery​ – more effective handoffs between job phases ensures a greater focus on outcomes

Solution landscape across industry

We offer unique set of digital solutions across each industry and domain sector

Click below solutions to see overview

Reliability & Predictive maintenance
Digital work management

Utility X

Vegetation X

Shape (PdM)

Wind AI

Mine X

ProVidens

ProdAI

CAMO

Prometheus

Here are some client examples

Large O&G player in SE Asia accomplished around ~$2m per year loss avoidance through ~30% reduction in unplanned downtime from pilot through McKinsey ProVidens

Click thumbnails to expand

Timely detection and diagnostic of bad diesel quality prevented fuel system replacement for 3 GTGs, leading to USD 1.5M+ impact

Click thumbnails to expand

Digitizing the maintenance process addresses key pain points along the maintenance cycle, leading to 3-5% equipment availability and 10-20% labour productivity improvement

Click thumbnails to expand

Latin America mining company eliminated overage, achieving 28% reduction in total execution time vs historical through Prometheus solution

Click thumbnails to expand

Deep offshore O&G operator reached unparalleled uptime for its global fleet through a unique system mostly based on predictive maintenance

Click thumbnails to expand

By working with us, you will have access to

Experts

Our industry and functional experts will be on the ground where and when you need them.

Helpdesk

Our helpdesk will keep doors open and commit to addressing your concerns efficiently and effectively.

Training

We offer coaching and build capabilities for your teams to deliver sustained improvements.

Implementation

We partner with you and co-create a firm path to implementation as part of the project.

Connect with us to find out more

© 1996-2023 McKinsey & Company