Back

Managing data quality with Artificial Intelligence Response Prioritization

October 09, 2024 | Written by: Harshkumar Patel, Jonathan Chong and Tiago Meira de Borba

Part III: Case Studies of Different Types of Data Quality Issues Considered by the AiRP Engine

Avalon Surveillance AiRP (AI Response Prioritization) is a state-of-the-art real-time surveillance system for ESPs (Electrical Submersible Pumps). Previously, Parts I and II of this series discussed several case studies showcasing how the AiRP engine proactively detects different ESP events, enabling timely corrective and mitigative actions and feedback to optimize operations.

Part III of this series presents case studies on various real-world data quality-related issues that are considered by the AiRP engine. Besides managing real-time expectations, automatically capturing data quality events enables a systematic evaluation of problematic data pipeline areas, leading toward continuous improvement efforts.
 
The AiRP engine is robust to data quality issues:

  • AiRP has a custom-built data quality (DQ) engine that identifies, in real-time, various issues with the streaming data (e.g., missing/outlier/frozen data, power loss, communication loss, etc.) and detects various electrical faults and gauge failures (e.g., a thermocouple failure or disconnected discharge pressure sub).
  • Based on real-time DQ flags, the engine can disable, enable, or modify the execution of various embedded machine-learning event detectors, adjust prediction confidence, and inform users of any compromised detection quality. For example, in the case of a disconnected discharge pressure sub-gauge failure, AiRP automatically handles all reliance on the discharge pressure signal for event detection and reports the impact of the compromised signal. If the drive frequency signal is temporarily missing, the engine automatically uses hydraulics and temperature signals to infer the status of the pump and motor.
  • More detailed discussions on the data quality engine and how AiRP achieves reliability and explainability are available in this paper.
Case studies:
The following are a few example case studies of data quality conditions that are considered by the AiRP engine.

3.1 Surface wellhead signals are missing 
3.2 Communication loss resulting in all channels missing
3.3 Broken discharge pressure sub gauge failure
3.4 Outlier data 
 
 
Case 3.1:
Surface wellhead signals are missing 
 
Description:
When one or more signals are missing, the AiRP engine raises data quality flags. Additionally, depending on which signals are missing, the engine automatically disables, enables, or modifies the execution of various underlying event detectors, adjusts prediction confidence, and reports detection quality status, as shown at the top of the screenshot below. Each bar represents an event, where green indicates the engine can raise it – while red indicates that the engine is not confident enough to raise that event, depending on the available signals and their data quality situation.
3-4.png
 
 
Case 3.2:
Communication Loss (CL) resulting in all channels missing
 
Description:
Due to a period of communication loss, all signals were missing and the AiRP engine appropriately raised the corresponding CL event (A).  For systems where the backfill of data is initiated upon the re-establishing of communications, these CL events are crucial for troubleshooting if the events have occurred in hindsight during the CL period, and the engine appears to have missed the detections of no fault of its own.  
3-3.png
 
 
Case 3.3: Broken discharge pressure sub gauge failure
 
Description:
This screenshot shows multiple types of DQ issues. The AiRP engine, having detected a gauge failure as the discharge pressure dropped and equalized close to the intake pressure, raised the appropriate disconnected discharge pressure sub-DQ flag/event. This means that the discharge pressure values are unreliable. Hence, the engine has reported an up-to-date detection quality and confidence through the AiRP status icon at the top left corner of the figure. Red bars indicate that, without reliable discharge pressure, the engine cannot confidently identify/distinguish certain events, such as Dead Head.
In addition, this poor DQ situation included frozen and missing signals that impact the engine performance.
3-2.png
  
 
Case 3.4: Outlier data
 
Description: This screenshot shows Outlier DQ flags. Intake temperature and intake pressure signals were frozen at 0 ºC and 18 PSI respectively. The AiRP engine correctly flagged both frozen and outlier DQ flags. This poor DQ situation also led to all AiRP status bars being red, meaning the engine is not confident to reliably raise any events. 
 
3-1.png

 
That brings us to the end of our AiRP Series. We hope you found it helpful in understanding how AiRP can support your operational needs. For more information on our AiRP products, click here. And remember to follow us on LinkedIn for more news, events, thought leadership on technologies that will help you achieve your objectives from the edge to the enterprise. For detailed information on all of our industry solutions, visit us at sensiaglobal.com 
 
Learn more about our authors:
 
Harshkumar Patel
As an R&D Data Scientist at Sensia, Harshkumar’s interest is in developing artificial intelligence and physics-based digital and automation solutions. He has over 10 years of experience conducting R&D across diverse domains within the oil & gas, geothermal, and offshore industries. Harsh (as he is known) holds a PhD in Petroleum Engineering from the University of Oklahoma, USA. He has published over 30 research papers and has received several scholarly awards, including the 'US Einstein Visa' for extraordinary research impact and international recognition in his field.
 
Jonathan Chong
Advanced Technology R&D Manager at Sensia, Jonathan is an expert in scalable operationalization of artificial intelligence and physical modeling techniques in digital and automation systems. He achieves increasingly autonomous operations that are robust and able to adapt to complex, evolving environments. Jonathan holds a PhD from the National University of Singapore (NUS) in the Innovation in Manufacturing Systems and Technology program under the Singapore-Massachusetts Institute of Technology (MIT) Alliance. He has 18 years of combined experience with SLB and Sensia in various roles within technology organizations, from conducting R&D and engineering to managing large technology programs.
 
Tiago Meira de Borba
With over 15 years of industry expertise, Tiago started a dynamic career primarily with SLB, navigating diverse oil environments ranging from offshore and deep-water subsea to onshore and unconventional settings. By 2023, he started a new role as Global ESP Domain Champion at Sensia. Here, Tiago is responsible for overseeing and optimizing artificial lift systems. He possesses in-depth knowledge of technologies like ESP (Electric Submersible Pump), Rod Pumps, Gas Lift, and other artificial lift methods. His primary role involves designing, managing, and supporting digital products and addressing technical challenges to achieve production needs.

Interested in reading more in the series? Click here to read:

Article 1 - How Artificial Intelligence Response Prioritization facilitates action-focused surveillance at scale

Article 2 - Enhanced decision-making and intervention enabled by Artificial Intelligence Response Prioritization

For more information please contact us

Your details
*required
Areas of interest (tick all that apply)
Please indicate your region
Enter security code:
 Security code