By Milad Mobarhan, Ph.D., Senior Software Engineer, Cognite
Many oil and gas companies are interested in exploring the potential of 3D, but are unsure of what to make of the technology. A complete 3D model of an installation may look impressive, but on its own, a model is no different than a map. And like a map, a 3D model can quickly become outdated unless it is continuously updated. So where’s the value proposition?
An example of how images can be turned into a 3D model (Source: Operator in the North Sea)
This is a common issue with computer-aided design (CAD) drawings of assets, which often seem to be obsolete as soon as an asset is put into production. Modifications or changes to the asset, like sagging under heavy loads, mean that companies must conduct manual measurements anyway. As a result, the CAD drawing simply can’t be trusted. No matter how detailed the drawing is, a static 3D model can’t capture the dynamic nature of operations at an installation.
Yet a static model is better than no model. In many cases, oil and gas companies simply don’t have CAD drawings of their assets.
Photogrammetry offers a solution to these challenges. By stitching together images to create photorealistic 3D models, oil and gas companies - and indeed companies in other heavy-asset industries - can create digital representations of the industrial reality at their installations.
While it may sound complicated, photogrammetry is actually a simple, low-cost technique for creating 3D models, and it can be accomplished by a single worker carrying a standard camera or operating a drone.
The process involves shooting high-quality images or video footage of the asset that is to be digitized, identifying and matching features that are visible across multiple images, and then using triangulation to turn the images into a 3D structure of the object. The images also contain texture information about the asset, which can be added to the finished 3D models to make them all the more photorealistic.
Once rendered, the 3D models can be made available across a company as a part of every worker’s toolkit. Uses include:
- Photorealistic 3D models help workers familiarize themselves with their surroundings, giving them a powerful, up-to-date tool for navigating installations and locating equipment.
- Companies can incorporate the 3D models into their planning work, virtually navigating the installation to determine the optimal path to a destination and seeing whether bulky equipment can be safely transported to the area where it is needed.
- The 3D models can be useful for inspection work. And as a company continues to capture new scans of its installations, engineers can study the different scans to see how the condition of equipment changes over time.
- The photorealistic 3D models can also complement CAD models, as the two can be compared to detect changes in an environment.
More Data, More Context
Photorealistic 3D models on their own already offer several features that can benefit oil and gas and other heavy-asset companies. However, the real potential of 3D is unlocked when it is combined with different technologies in order to make data more accessible. Live sensor data is an obvious candidate. By using computer vision to read all the asset tags visible in the images used to create a 3D model, a company can link all the data it collects about the different equipment captured by the model to their digital counterparts. The result is a photorealistic digital twin that reveals new insights about previously separate data sources.
Imagine the wealth of use cases for this kind of technology in the oil and gas industry. For example, picture a 3D model of the engine room on a floating production storage and offloading (FPSO) unit, in which each component is clearly labeled with live-streamed data about pressure, temperature, and more. As long as the 3D model is kept up to date and the data collected is of sufficiently high quality, inspecting the engine room could be as simple as pulling up the model on a computer or smart device.
The “up-to-date” factor is currently one of the major obstacles keeping oil and gas companies from exploring photogrammetry. Having a worker walk around an installation and capture hundreds of pictures every day can be a full-time job in and of itself.
This is where automation can help. Imagine a future in which autonomous drones and robots conduct daily photography rounds, navigating every nook and cranny of an installation to continuously update the 3D models. Now imagine applying computer vision to those models to detect corrosion or a faulty cable, automatically adding the potential issues to a work order and notifying engineers.
That future isn’t far off. Photogrammetry can be a piece of it, one part of a larger industrial digitalization strategy that aims to liberate data from its silos, connect it to other relevant information, and make it available to the workers who need it.
Oil and gas companies can already now get a sneak peek of the digitalized future. All they need to do at this stage is grab a camera and head out into the field.
KeyFacts Energy Industry Directory: Cognite