What is the Digital Future?

If the recent downturn has taught us anything, it is that we must, and can, become more efficient in operations. Becoming more efficient means either consuming fewer resources (time, money, equipment, people) for an equivalent result, or consuming equivalent resources for an improved result. A more efficient operation delivers better results, improving the top line, and reduced costs, improving the bottom line.

The keys to improving and ultimately optimizing an operation’s efficiency are to use improved data, information, and knowledge about the way the operation consumes its resources.  This improved data/information/knowledge flow allows us to make better decisions about planning, operations, and maintenance of our assets. Once we achieve an optimum efficiency, we can move on to becoming more effective, but that’s for a later discussion.

The “digital future” implements measurement, control and inspection technology, improving the quality and value of the data/information/knowledge produced during operations. This improved data/information/knowledge flow can increase the output and lower the cost of operations by enabling better decision making for both real time and planning activities. This process, whether applied to individual assets, fleets or organizations looks like this[1]:

[1] Referenced from http://www.tlainc.com/articl134.htm

The digital future allows this data/information/knowledge flow to be used in a distributed framework, utilizing subject matter experts that are both local and remote to the actual activity, ultimately reducing or even eliminating resources required to run that operation (for example, the number of local persons (POB) necessary).

The operational activities best positioned to benefit from a digital future include:

  1. Pro-active (predictive, preventative, risk based, reliability based) rather than reactive (break-fix, schedule) maintenance and inspection programs
  2. Executing efficient and effective operations that consistently track within the acceptable performance bands
  3. Operational excellence and improvement programs
  4. Remote assistance and diagnostics
  5. Mechanization and automation
  6. Systems integration

The Digital Future starts with the Data

Consider as an example, the first of the six operational activities listed above. Similar to the rush to automation, one needs to be very careful about the rush to use operational data (alarms and warnings logs for example) to make predictions about future equipment operations. A predictive analysis is a final step, not a first step. The first step is ensuring the data integrity.

Before you can analyze a data stream to make a prediction about an equipment-related event which could cause downtime, you must first ensure that the data stream in fact contains the necessary data. For example, a large majority of data produced on an operational rig is sampled, and the nature of sampled data means that critical information can and will be missed if it is not sampled correctly. The aliasing of an under-sampled digitized waveform is a prime example of this.

The well-proven and longstanding basic principles of equipment control establish that three steps be accomplished before you can use data from that equipment to generate operational or predictive analysis.

  1. Ensure you can make the correct measurement – this means that you must first fault model the equipment and, identify the faults that are critical to operations (HAZID, Risk Analysis, FMECA). Once the critical fault modes are identified, you must identify the equipment indicator that would signal that fault has or will occur. Finally, you must define a sensor and a measurement (sensor analysis) that can detect and report that indicator. You cannot analyze what you cannot measure.
  1. Ensure that the measurement is stable and reliable – Once the measurement has been defined, you must ensure that the data being returned represents a statistically normal distribution for that process or equipment. One cannot make statistically based decisions (the root of predictive analysis) on data that is not stable. If the data is not stable, the process or equipment must be stabilized before its performance can be predicted.
  1. Characterize the equipment – Characterization is the final critical step when determining if you can in fact make predictions about future events. Characterization involves analysis of past performance to find the patterns and behaviors that allow predictions. If historical data does not indicate that there are predictive behaviors, then one cannot trust the data to make future predictions.

Athens Group provides the full suite of services for HAZID, Risk Analysis, Fault and Failure Modes analysis, sensor and measurement integrity and equipment characterization necessary to ensure your data can provide the predictive analysis you need to ensure continued high uptime operations. Contact us at info@athensgroup.com for more information about taking the first critical data integrity steps towards the digital future.

Copyright 2017 Athens Group.
Subscribe to our Newsletter

Subscribe to our Newsletter

* indicates required