As the digital transformation of the energy industry continues apace, organizations are now more focused than ever to get the most from their investments in both assets and software. Asset Performance Management (APM), a foundational solution for operators, has been a growing and evolving market that taps into this need to optimize processes and increase return on investment (ROI). Along with increasing digital capabilities, operators are also making the shift from reactive to predictive maintenance (PdM) as a way to reduce operations and maintenance (O&M) spend and risk.
As they make the change to a predictive operation, organizations should raise the question: how healthy is the data being used to monitor asset conditions? At GE Digital, we’ve identified five levels of equipment health monitoring powered by analytics, each of which depend on healthy sensor data:
Unfortunately, unhealthy or “bad” data is a fact of life. Sensors fail, and connectivity can occasionally be lost. Such failures most often manifest as missing, flatline, erratic, or extreme data values. Sometimes sensor health is restored quickly, while at other times it may take weeks or even months before equipment can be taken offline for repairs to be made.
Sensor and connectivity failures need to be addressed as quickly as possible if Digital Twins, which apply advanced analytics and machine learning to a virtual representation of the equipment, are to continue functioning optimally. The analytics themselves can play a part in dealing with unhealthy data.
For these reasons, GE Digital has been working with our customers to develop Sensor Health Monitoring (SHM) as part of our APM Reliability solution that focuses on powerful predictive analytics. By identifying and improving any gaps in sensor data, users can have more confidence in alerts from the application.
Sensor Health Monitoring (SHM) employs intelligent online algorithms that continually monitor data feeding into analytics. It provides a notification in the form of an alert, as well as automatic maintenance actions, when analytics experience unhealthy data inputs. SHM is designed to help organizations more accurately detect equipment faults by identifying bad tags and prioritize when and which actions are taken.
In SHM there is at most one alert notification per asset, which is updated continually and includes:
SHM also helps to reduce alerts that could lead to unnecessary maintenance actions and helps reduce the need for user intervention. Some examples of maintenance action reduction include:
As SHM actively detects and responds to bad sensor data, the parts of the analytic that have not been directly impacted by unhealthy data will continue to function as intended.
With SHM enabled in APM Reliability, users will also gain access to historical records of sensor health activity, an analytic-level view of unhealthy data inputs, as well as their recovery over time. Through this capability users benefit from more visibility into overall sensor health and the impact within the operation.
SHM allows organizations to get the most from their investment in APM Reliability. Through more accurate alerting, Reliability teams can prioritize and reduce time dispositioning alerts, reduce the number of unnecessary alerts stemming from bad data, and act faster on priority sensor alerts. It automates some sensor evaluation and empowers engineers to spend their time on higher value activities.
SHM is available today, at no additional cost, for current APM Reliability customers and will be included in any new APM Reliability purchase from GE Digital. Contact a GE Digital expert to learn more about how SHM can bolster your reliability program.