Pioneering Oil & Gas Enterprise

Case Study

Transforming Operations Through AI-Driven Solutions

The client, a leading mid-stream Oil & Gas company with a large network of pipeline infrastructure faced significant operational challenges, particularly in managing alarm systems and ensuring effective prioritization of client needs.

‍Recognizing the potential of artificial intelligence to enhance their operational efficiency and address these issues, the company partnered with Reyman,

‍The collaboration started with an extensive phased proposal designed to harness AI for optimizing the client’s operations. As project discussions progressed, the Reyman team focused on two primary initiatives: the Alarms Management Platform and the Client Prioritization System.

Nearshore Solutions

AI / ML
DevOps
Custom Software

Development Expertise

Python Development

Outsourced Services

DevOps and Cloud
Software Development
AI and ML Development

The Challenge

The client confronted a critical challenge with a high volume of false alarms. These frequent disruptions inundated the control center, complicating the prioritization of genuine issues that required immediate attention.

‍To tackle this, the client needed a solution that could effectively identify and filter out false alarms and significantly reduce noise in the control center.

‍Additionally, they required a tool to monitor when a customer exceeded their permitted usage limits, which could potentially block other customers from accessing the system. These challenges highlighted the need for a robust system to enhance operational efficiency and improve communication.

The Solution

Reyman designed an innovative Alarms Management Platform with several key components:

Alarms Overview Section: This section features an initial dashboard summarizing vital alarm information. It incorporates filtering capabilities, allowing operators to adjust the displayed information based on specific criteria. A heatmap visualization further empowers the control center by illustrating alarm behavior throughout the day, making it easier to identify patterns and trends.

Alarms Overview Section: An initial dashboard summarizing key alarm information, with filtering options to customize data views. The heatmap visualizes alarm behavior throughout the day.

Alarms Details Section: This primary panel provides real-time updates on alarm events, with a specific grouping of alarms based on the client’s definitions of duplicated events and nuisance alarms. The integration of a “False Positive” column enables the AI model to detect and classify false alarms, while also allowing client feedback on any missed cases. This feedback loop facilitates continuous improvement of the model during retraining processes.

Alarms Details Section: A primary panel displaying real-time alarm events, grouped according to the client’s definitions of duplicated and nuisance alarms. The “False Positive” column shows alerts flagged by our AI model, allowing clients to review and provide feedback for ongoing model improvement.

AI-Powered Features: Reyman employed an anomaly detection model using techniques like Isolation Forest and Autoencoders, trained on sensor data related to specific alarm events at designated sites. For example, by establishing ground truth using high-pressure situations as labels, the model matches triggered alarms with corresponding sensor data. Feedback from users about false positives further refines the model, ensuring it accurately distinguishes between true alarms and nuisance triggers.

***Real-time Alerts Section:*** *The Notification Log displays SMS messages sent in real-time when specific situations occur. The Alerts panel below provides detailed information about the events that triggered each SMS.*

Additional Functionality: To address the need for monitoring resource usage, Reyman implemented a feature that sends automatic notifications via SMS whenever a customer approaches their usage thresholds. This proactive communication enables timely actions to prevent operational bottlenecks.
Real-time Alerts Section: This section features a Notification Log that displays SMS messages sent through the system, providing immediate visibility of critical situations. The Alerts panel offers detailed insights into the circumstances that triggered the notifications, ensuring the client is always informed.
Inbound Notifications: The system will accommodate maintenance notifications from field teams. When maintenance is scheduled, relevant alarms will be filtered into a separate table, minimizing clutter in the main dashboard and allowing operators to concentrate on real-time issues.
Outbound Notifications: To enhance communication, the platform will introduce a notification button that automatically informs field teams of relevant alarm data and messages, streamlining communication and response times.

Getting Results

Through this collaborative effort, Reyman has significantly advanced the client’s operational capabilities. By reducing false alarms and enhancing resource allocation, the partnership improves efficiency and positions the client for sustained growth in a rapidly evolving market landscape.

The ongoing commitment to innovation and responsiveness to client needs exemplifies Reyman’s dedication to delivering impactful solutions that drive success.

The Client Prioritization System aims to leverage AI for optimizing resource allocation, and ensuring adherence to client contracts while maximizing revenue. Future enhancements include evaluating alternative processing locations and more.