Reimagine Pipeline Inspections with These Aerial Data Products

A digital elevation model of a pipeline right-of-way

Pipelines across the United States are getting old. According to reports by the Pipeline and Hazardous Materials Safety Administration (PHMSA), approximately 55% of gas transmission lines and 43% of hazardous liquid pipelines are over 50 years old – that’s over half a century!1 As assets age, they become more susceptible to hazards. However, with rigorous inspection programs and preventative maintenance, you can safely prolong the life of your existing infrastructure or make a well-informed decision to replace a segment.

Unmanned aerial data platforms can serve as a vital tool in your asset integrity arsenal to meet this challenge. Through largely-automated processes, these platforms can deliver models and precise visual insights that revolutionize how you monitor your pipeline and Right-Of-Way (ROW).

Computer Vision Analytics: Pinpoint Failures and Hazards

While Unmanned Aerial Vehicles (UAVs) offer imagery with fantastic spatial resolution and accuracy – this is only the raw input for valuable aerial data. Every inspection of your ROW is going to generate thousands of images for painstaking analysis. Wouldn’t it be nice if you could get proactive anomaly detection that your team can take action on with each inspection? Well, you can! Through computer vision technology.

At SkyX, we utilize deep neural network models and algorithms for computer vision analysis. After each inspection, an anomaly detector that is trained to recognize midstream pipeline hazards on a right-of-way meticulously combs through each and every photo. Pooling liquids, dying vegetation, third-party activity, and other environmental factors – the anomaly detector can flag them all. Each and every inspection can be viewed as its own dataset. Over time, through change detection analysis, we can compare new and historical datasets and notify you of any noteworthy changes happening to your right-of-way – be it consistent erosion or slowly-encroaching construction.

The summaries and coordinates of these anomalies provide massive value as annotations in your GIS, but they’re also extremely useful when incorporated into other aerial data models.

Orthomosaics: a Digital Record of Your Pipeline Right-of-Way

Orthomosaics introduce a new level of visual acuity to the GIS interface you use to manage your operation. An orthomosaic is a large, map-quality image made by combining the individual photos taken on a right-of-way, providing high detail and resolution on a map-wide scale.

In your pipeline operations, an ortho serves as an updated digital record of your ROW. Oftentimes, GIS basemaps are based on outdated satellite data, which offers poor resolution from an inspection standpoint. With a UAV-sourced orthomosaic, you can zoom into an area of concern without substantial loss of resolution.

Having this kind of visual record enables greater compliance with regulators and external stakeholders. With the ability to provide a high-resolution visualization of an area, along with geospatially-accurate coordinates to overlay important operational and property data – both parties can work from a place of common understanding.

An ortho also gives you a better understanding of the ground conditions around your operational overlays in a GIS. When you combine computer vision annotations of potential issues on your ROW, you create a singular interface from which you can effectively manage your entire operation.

Digital Elevation Models: Understand the Surrounding Terrain

The numbers behind an image can also be of great value to an operator, and a Digital Elevation Model (DEM) provides a mathematical representation of your site’s ground surface elevations. Traditionally, these models required expensive satellite surveys and highly specialized radar or LiDAR sensors. Today, these models can be created easily through the combination of a simple UAV-mounted camera and photogrammetric mapping techniques.

From a DEM, a pipeline operator can analyze key environmental factors and how they may affect operations:

  • Track the rate of erosion or ground movement to understand the amount of stress being put on the pipeline
  • Perform hydrological analysis to model water crossings that intersect your ROW, drainage, and other catchment areas for analysis.

Again, this is another instance where combining different forms of aerial data enables a holistic analysis of your ROW. When computer vision analysis flags a potential ground movement anomaly, such as a crack, sinkhole, or severe instance of erosion – you can cross-reference that anomaly against the elevation readings from your DEM to understand the true extent of the ground loss.

3D Point Clouds: Precise 3D Modelling of Your ROW

If only the utmost accuracy will do for your pipeline ROW survey, a UAV-mounted LiDAR sensor can be used in the creation of 3D point clouds. Point cloud data is useful for the creation of laser-accurate models of the ground surface and terrain, especially given the fact they’re able to capture readings of the ground below dense tree canopies. Additionally, with their 3D spatial data, a point cloud is very useful for modeling structures or even the vegetation around your right-of-way in 3D space. For example, in a scenario where a segment of the pipeline is enduring extreme flooding, a LiDAR survey can help track how the asset and ROW are being affected by these conditions.

Keep in mind, you can introduce other specialized sensors to your UAV remote sensing program data over these models to better monitor for specific issues along your ROW.

Focus on Key Pipeline Inspection Data Through AI and Predictive Analytics

All of these data deliverables give you important real-time insights, but AI can help you transform them into predictive analytics. The power of an aerial data platform is it translates what you see visually into quantifiable data. From quantifiable data, we can identify trends and make forecasts for what to expect along your pipeline. When an actionable issue is identified, the system logs all the relevant information – the nature and severity of the issue, coordinates, time/date – and creates a critical point of data that contributes toward a predictive model.

Integrating other important sensory data from your pipeline monitoring program supercharges the insights you can obtain through predictive analytics – such as readings from your SCADA point sensors, in-line PIG inspections, or even external data sources like weather conditions. Say the aerial platform identifies dead vegetation, which reveals a previously undetected pinhole leak. Cross-referencing this incident with the most recent PIG inspection can help build a profile of when your pipeline is approaching a state of instability. Another example, In the instance of SCADA alerts (which struggle with localization), an aerial inspection can help pinpoint where the failure is occurring, allowing you to identify the problem areas along that segment.

In practice, predictive analytics help you make the best decisions to manage an aging pipeline:

  • Problem areas where preventative maintenance efforts should be allocated
  • The profile of developing issues along your line, allowing you to flag them before they occur
  • How small changes are gradually leading to bigger issues, such as the erosion of earth above your pipeline over time or slowly-encroaching excavation activity

Managing an aging pipeline safely and responsibly requires proactive integrity management. Through aerial data platforms, you can obtain the predictive insights that empower you to make intelligent decisions around preventative maintenance that prolong the life of your asset. Or when the data tells you it’s time to make the hard call of replacing the pipe.

Have questions about how high-quality aerial data can elevate your organization?
Contact our team to discuss your unique challenges and data requirements.

References

  1. Miles by Decade Report Pipeline Hazardous Materials Safety Administration, 2020