iotSymphony Use Case - Predictive
Maintenance for Oilfield
Services

  • user

    100+ Years of Enterprise
    Software Experience

  • gcp logo

    GCP

  • concept

    No-cost
    Proof of Concept

sec-bottom-line

Business Challenge:

The global energy crisis has created increased demand on all types of fuel, including fossil fuels. Whether oil drilling, ore mining, natural gas drilling, or any other extraction process, fossil fuels remain a critical source of energy over the coming decades until there is widespread adoption of alternative energy sources. Oilfield Services encompasses many of these fuel extraction processes, including natural gas drilling, oil drilling, and ore mining, as well as their associated support services.

Each extraction process requires multiple heavy equipment components, including shale shakers, mud pumps, fracking pumps, blowout preventers, hydration units, shotcrete machines, and many more. Any time one of these heavy machines fail to operate creates tremendous production risk and cost implications to oilfield services companies. Therefore, Predictive Maintenance is a critical capability, and one that requires the Internet of Things.

Most heavy equipment contains multiple sensors that capture data on the machines production performance as well as its operational performance, including temperature, vibration, pressure, speed, and many more. All this data needs to be monitored, often at very short time intervals (per second, sub-second), resulting in huge volumes of streaming data that must be assessed by oilfield services operations teams. Solely relying on human assessment of this data results in missed performance degradation indicators, which create multiple challenges for oilfield operations, including:

business-challange
  • Missed operational performance conditions that are indicative of impending machine failure
  • Numerous “false positives” that inhibit the ability for operations teams to identify performance anomalies
  • Over-worked field operations teams as time is spent extracting, compiling and assessing machine performance data

Impacts to the business:

High frequency of machine down time due to missed opportunities in accurately identifying performance degradation issues, extensive man-hours spent by maintenance technicians investigating false positive performance alerts, and an inability for oilfield services companies to maximize the value of all the data their operations generate, all of which reduces productivity, increases operations costs, and increases liabilities as field teams are erroneously dispatched to perform unnecessary machine inspections. As oilfield operations teams work to overcome these issues they have brought on more staff which further increases operations costs while still not achieving the values they are striving for.

heavy-machine
better-way

A better way with iotSymphony:

By leveraging the iotSymphony software platform, oilfield operations can dramatically improve their predictive maintenance capabilities by utilizing complex alerts, AI/ML and decision science to assess machine performance in real-time, identify performance degradation issues and automatically dispatch the right maintenance technician with the appropriate skills and tools needed to resolve the issue before catastrophic failure occurs. And here is how:

  • As telemetry data streams from heavy equipment, iotSymphony captures this data, applies operations analytics and runs it through advanced business rules to determine if the data requires further analysis
    • Advanced business rules take into account variations in performance data and how each data element performs in relation to other data elements
    • Advanced business rules can be prioritized by rule and rule set to aid in reducing unnecessary alerts and ensure the most impactful performance data is considered first
    • Based on the alert, a decision flow can be activated that aligns to enterprise business objectives. This decision can automatically activate the required action in the operations system, such as dispatching a maintenance technician with all necessary skills, tools and parts to expediate repairs
  • If the alert generated requires additional analysis, the data can be merged with other pertinent data (manufacturing data, location data, weather data, etc.) and then processed by the analytic engine where AI and ML models are executed
  • Based on the outcome of the analytic processing, that data can be run through a decision flow to determine the best course of action
    • Decisions are aligned with enterprise business objectives
    • Decisions can be extensive, such as replacing a machine, or moderate, such as replacing a part, or simple, such as do nothing and monitor
    • There is no limit to the number of decisions that can be created
  • Based on the outcome of the decision flow, an action can be automatically generated
    • This action can be to automatically dispatch a maintenance technician
    • This action can be to notify the data science team to perform additional analysis
    • The action can be to generate an action in an operational system (any system)
    • The action can be to simply send data to the reporting portal
    • The action can be to simply do nothing
    • There is no limit to the number of actions that can be activated

Outcomes for the business:

Machine downtime rates have significantly dropped due to having effective predictive maintenance procedures leveraging automation firmly in place. The utilization of operations teams has improved as well as overall operations efficiency, resulting in a reduction in operations costs. This increases profitability from oilfield operations and reductions insurance liabilities as the number of technician dispatches decreases with a reduction in false alerts.

outcome-business