Visualization and Simulation in industrial operations

This report reviews how the industry brings reality to the screen and how real world operations and processes can be replicated (simulation) to help field operators perform their tasks with greater accuracy and efficiency.

In this report we will review:

  • Visualization Technologies. Pros and Cons.
  • Visualization Technologies and Simulation.

VISUALIZATION TECHNOLOGIES. PROS & CONS.

Some technologies replicate reality exactly as the human eye sees it, although by and large, a visualization technology is any technique used to create a visual representation and to communicate a message.

In this article we will analyze:

  • Photographs
  • Spherical Panoramas
  • Video / Spherical Videos
  • Computer Generated 3D Model
  • Enhanced Reality

PHOTOGRAPHS

A photograph is a durable reproduction of reality that the human eye sees at a specific moment in time. The viewer cannot interact with this representation of reality. Photographs are a means to communicate a moment in time or place to a third party.

Image 1. Photograph

Nowadays technology has lowered the barrier to taking good photographs. Anybody can take a picture with their phone, point and shoot camera or SLR, and share it.

SPHERICAL PANORAMAS

A Spherical Panorama (SP) is a full-view photograph made from a fixed location. The viewer/user can look up, down, left, right and behind. SPs are commonly used to share a location.

An SP is a single photograph created by stitching several photographs together that have been taken from a fixed location. The sum of all photographs covers the full-view of the location (Image 2), and a player is used to let the viewer/user look around and up & down.

Image 2. Spherical Panorama

Because an SP is the result of stitching several photographs (it can range anywhere from two to thousands, depending on the desired resolution), its final size may impact its potential uses.

The most common ways to create SPs are:

Taking photographs with a normal camera mounted in a “pano head” (to avoid the parallax effect [1]), and stitching them with special software.

Image 3. Pano head

Using spherical 3D cameras. These cameras are usually ball-shaped with several lenses, processing power and wifi capabilities. Some of these cameras can create SPs directly without the need of further processing.

Image 4. Spherical 3D cameras

VIDEO / 360º VIDEO

A video is a series of photographs taken sequentially in such a way that when they are interpreted by a player (video player), it replicates reality in a timeframe. Videos are commonly used to tell a story and the viewer/user is typically a passive spectator.

If each photograph in the sequence is a spherical panorama, the video is called 360º video. In this case the viewer/user can create part of the experience because she or he decides where to look from their viewpoint while watching the video (remember the degrees of freedom in an SP).

Nowadays anybody can shoot a video and share it, i.e. with a smartphone, but there are two major limitations:

  • It is not easy to record a good quality video with a moving camera.
  • Resulting file sizes can be large depending on factors like resolution, frame rate and length of the video.

Major setbacks of 360º videos are:

  • Special equipment is required (spherical 3D cameras or a rig of cameras) as well as professional shooting.
  • File sizes can be large. As a rule of thumb a 360º video file would be 9 times larger than its equivalent Full HD video file [2]
  • Special players are needed to view and interact with the 360º video.
  • Unless you are using a high-end spherical 3D camera, it is very difficult to create a good quality high-resolution 360º video as editing is required.

COMPUTER GENERATED 3D MODEL

Computer Generated Models, also called CAD Models, are navigable artificial 3D models designed with a computer system. The viewer/user can walk in any direction and look up, down, left, right or behind, and can also interact with some elements in the 3D model. CAD models are used for design and manufacturing purposes but also for simulation and video games.

Although the underlying constituents of the 3D model are vertices connected by straight lines or curves (wireframe), a 3D image (also called 3D scene) can be created via 3D rendering:

Image 5. Evolution from a 3D wireframe to a 3D rendering

As an example, two rendered images of different quality and resource requirements are shown below.

Image 6. Schematic Appearance — Low quality rendering & fewer resources needed
Image 7. Photorealistic Appearance — High quality rendering & more resources needed

The level of realism affects the credibility and value of the rendering for simulation purposes, and the rendering quality impacts other factors. Increasing the quality of the 3D scene rendering implies:

  • File size increases.
  • Updating the 3D model becomes very difficult and hence expensive.
  • Technical knowledge requirements of the rendering team increases.
  • Production time increases.
  • Processing power of hardware to use the 3D scene increases.

One of the biggest setbacks with 3D scenes (already rendered) is the time and amount of resources needed to update them. The process of updating means, among other considerations, altering the wireframe with new elements or changes, running all the needed simulations, creating the surfaces and defining their material and the 3D rendering itself. Therefore updating CAD-based scenarios is a very resource intensive process.

ENHANCED REALITY

Enhanced Reality (ER) creates a photorealistic 3D scene from a collection of photographs. The viewer/user can move around continuously with no interruption and look left, right, and behind and can interact with elements in the 3D scene. ER is used to replicate and share a big place, building or asset, as well as for the simulation of how a worker interacts with industrial equipment in a facility or large vehicle.

ER is a technology that transforms a simple still photograph into a 3D self-contained navigable space where a user can move freely. With a series of photographs ER creates a continuous 3D scene as every single 3D space of each photograph is linked to the others via 3D navigable connections.

Image 8. ER interactive scenario

ER technology uses a set of photographs taken along a desired path as its raw material to build interactive scenarios. The equipment needed for photoshooting is a normal camera, and the size of the files produced is much lower when compared to a video covering the same area. For illustration purposes, given a 70 meter walk:

  • A video of the walk will have 3000 photographs (assuming it takes 1 minute to walk 70 meters and the video is shot at 50 frames per second)
  • An Enhanced Reality version of the walk will have 14 photographs (ER needs an average 1 photograph taken every 5 meters)

So, with no compression, a video will roughly be 200 times larger in size than the equivalent ER version as the proprietary files added by ER are also negligible in size.

The nature of the 3D scene (made up of discrete photographs connected by virtual links) makes updating straightforward because it is just a matter of changing old photographs for new ones. No special equipment or high production time is needed.

Enhanced Reality 3D scenes are easy to share because they are a light solution and the player required to view the scenes is coded in standard javascript, which means the 3D scene can be streamed and watched in any HTML5 browser.

VISUALIZATION TECHNOLOGIES AND SIMULATION

In this section we analyze how suitable different visualization technologies are for simulation purposes. In particular we are interested in simulations of real world operations in industrial facilities, with the goal of helping field operators train and execute operations and maintenance procedures flawlessly and efficiently.

This specific simulation requires that the visualization technology (scene capabilities) excel in:

  • Realism: It is critical that the scenario be photorealistic to ensure credibility, user immersion and engagement. Any user needs to perceive what she or he sees on the screen as ‘real’.
  • Degrees of freedom: For user engagement and credibility, the scene needs to support the same degrees of freedom as in real life. In other words, it needs to support continuous movement (when walking from point A to point B) and capabilities of looking around (when moving the head (view) around in a fixed location).
  • Continuous navigation: User engagement requires the experience to be seamless when there is movement or interaction.
  • Interactivity: The scenario needs to support interaction with elements. These elements need to behave as in real life, and replicate the logics of action-reaction.
  • Easy updating capabilities: Plants and processes evolve on a day-to-day basis, so any technology that makes it difficult or expensive to update elements in the scenario is doomed for failure.
  • Mobile: Simulation needs to run on mobile devices, via streaming or standalone through a web-browser in order to facilitate the training opportunities and to accompany the user in the planning and execution phases if needed.
  • Production time: The shorter, the better.
  • Accessible technology & easy implementation: The more complex it is to understand and implement a solution in an organization, the more barriers this technology will encounter to succeed.
  • Cost: the return on investment must make sense.

This table, compares the visualization technologies as per these parameters:

Image 9. Visualization technologies comparison

[1] Is a displacement in the apparent position of an object viewed along two different lines of sight[2] As an example a Full HD video (resolution of 1920 x 1080 px) has 9 times less resolution than a 360º 4K video (resolution of 4096 x 2048 px). On top of it, if we take into account a normal FOV of 120 degrees, the end resolution in a 360º video will be smaller than in the equivalent flat video.


Originally published at voovio-blog.webflow.io.