2020 AFPM Summit: Predictive maintenance: A digital transformation success story

Predictive maintenance: A digital transformation success story



Given the current state of the global economy, it has never been more important to reduce costs and improve operations. Predictive maintenance using innovative operational artificial intelligence (AI) technology has emerged as one of the most easy-to-use and cost-effective ways to accomplish this. This article details the story of one large marine energy transportation, storage and production company that was successful in their digital transformation journey using operational AI software to improve predictive maintenance, resulting in substantial financial and time savings.

Compressors: Mission critical equipment. The company’s offshore oil and gas operations were highly dependent on rotating machinery. Floating vessels—such as floating production storage and offloading (FPSO), floating liquid natural gas (FLNG) and floating storage and regasification units (FSRUs) —rely on such compressors for their operations. It was difficult for this company to predict FPSO and FSRU failures that were resulting in production loss.

On average, unscheduled downtime for the company was resulting in 36 hr of lost production at $200,000/d, or $300,000 per incident. In addition, their modern LNG tankers relied on compressed fuel gas for propulsion. A compressor shutdown would force the tanker to burn the company’s cargo until it could be repaired. When this happened, two costs were incurred: loss of daily rate from the cargo owner, and penalties for consumption of client cargo that could be as high as $1 MM.

Applying predictive operational excellence. Using operational AI software from Falkonry, the company’s engineers were able to rapidly create predictive models that automatically discovered patterns in their time series data and enabled them to identify precursor events to the compressor failures. The engineers used the existing historical data with known failures to learn a predictive model. Some of the signals included in the modelling of the Burckhardt Laby-GI compressor on the LNG vessel were compressor motor current, suction pressure, differential pressure, discharge pressure, suction temperature and discharge temperature.

The high-pressure compressor was equipped with a Prognost-NT monitoring system. As shown in FIG. 1, the LNG carrier model was trained to find the precursor condition and predict the impending valve failure 6 wk in advance.

FIG. 1. The LNG carrier model was trained to find the precursor condition and predict the impending valve failure.

Results. Using Falkonry operational AI software, it took the engineers only 20 min to start automatically discovering patterns that, in turn, could be used to identify precursors to the valve failure. Further, the company was able to quickly move from initial trial to a full production deployment of predictive operations across multiple compressors and vessels, resulting in significant operational improvement and maintenance costs savings.

The deployment enabled them to predict compressor seal failures 6 wk in advance, much earlier than internal operations teams would find suspected issues. The total benefit from the initial pilot of two FPSOs was estimated to be $580,000, with the primary benefit being downtime reduction. Additional benefits included reduction in installation cost and increased compressor life.

Predictive maintenance: The next big opportunity. Operational AI technologies that help companies avoid unexpected downtimes represent a major competitive opportunity for oil and gas companies. This digital transformation journey can be easy to begin, with rapid results that help companies save valuable time and money when working with the right technology. Before embarking on your own journey, review these helpful hints so you are best prepared for success:

Choose a predictive maintenance solution that is easy to use by your own internal operations engineers.

  • Make sure the solution you choose can deliver results quickly.
  • Use a solution that can be applied to address many use cases and operation problems
  • Ensure it can be easily integrated with your data architecture to quickly deploy and operationalize in your environment.

For more information, visit www.falkonry.com


NIKUNJ MEHTA founded Falkonry after realizing that very valuable operational data produced in industrial infrastructure goes mostly unutilized in the energy, manufacturing and transportation sectors. He believes hard business problems can be solved by combining machine-learning, user-oriented design and partnerships.

Prior to Falkonry, Dr. Mehta led software architecture and customer success for C3 IoT. Earlier, he led innovation teams at Oracle focused on database technology and led the creation of the IndexedDB standard for databases embedded inside all modern browsers.

He earned a BS degree in computer engineering from the University of Mumbai, and holds an MS degree and PhD in computer science from the University of Southern California.  He has contributed to standards at both W3C and IETF, and is a member of the ACM.

From the Archive



{{ error }}
{{ comment.comment.Name }} • {{ comment.timeAgo }}
{{ comment.comment.Text }}