4 ms·

There's still a huge pursuit around this in most of manufacturing. Operational engineering groups are pushing for high fidelity digital twins to test out proces

by vnxli 6y ago

There's still a huge pursuit around this in most of manufacturing. Operational engineering groups are pushing for high fidelity digital twins to test out process control code before it hits the physical plant. The merge between simulation games and simulations gets even more blurred now that some software companies are replicating the entire plant in a 3d "game engine" and using VR goggles for immersion for operator training

KineticLensman 6y ago
A company I used to work for built a large sim for an Australian mining company, using a commercial game engine and a large 'igloo' type structure as the interaction environment. This was chosen over VR because it was believed to be easier to have groups of trainees in the same environment. The trainees were equipped with physical 'torches' that were geometrically paired with the virtual environment so you could actually shine a pointer on objects. The main training tasks were spatial orientation in the large 3D maze that was the mine, as well as hazard recognition and safety /;emergency training. E.g. knowing which way to go if an explosion occurred and how to avoid being crushed by the large vehicles that moved through the mine. The main unexpected development cost turned out to be the high degree of customisation required of the virtual assets, in particular the signage. Every tunnel and shaft had individual location and safety info on signs, and these had to be correct or the training value was lost.
- vnxli 6y ago
  The high-fidelity replication is much easier for things like industrial control systems (ICS)because all that needs to be replicated is the operator interfaces. Those interfaces are usually intentionally bland and typically dated - meaning really simple, standardized graphical assets I can imagine a complete mine recreation would be a huge development effort - especially with how huge they get. Is that something that like a camera drone could scan? i guess it could get the topology figured out with some special software, but could it get the signs and landmarks too? Can drones fly effectively in a mine?
  - KineticLensman 6y ago
    > but could it get the signs and landmarks too? IIRC the customers supplied the primary graphics content for the signs digitally from their own files; the underestimated development cost was finding and editing the appropriate instances of the signpost objects. The corridor rock walls were a generic texture. There were also lots of physical 'emergency exit this way' indicators (e.g. arrow shapes that could be felt by touch in darkness) on railings. These were instanced as well, but the direction had to be checked. > Can drones fly effectively in a mine? No idea but think lots of corridors with short Line Of Sight. Probably easier now to use Lidar or similar. [Edit] Also, some of the signs were dynamic, e.g. "Checked by X on date Y" and needed to be updated in a scenario-specific way.
  - kortex 6y ago
    > Those interfaces are usually intentionally bland and typically dated - meaning really simple, standardized graphical assets This sounds kind of like "worse is better". it got me thinking about user interactions, both gui and physical objects, from the angle of "what if I had to make these as 3D assets?" Screens suffer from limited real estate. But what could eg a CI pipeline dash look like in VR? You could have a whole virtual "control board" like a power plant.
  - ptrott2017 6y ago
    Re: Drones - not in the sense I think you are imagining, but you can pull data from some of the mining equipment and I have seen 3D scanners mounted in situ at test sites to capture specifics. (it didnt work well - got remodelled later). The general workflow I have seen would be familiar to anyone in game or animation pipelines. A team models from the mine layout map in traditional 3D tools (e.g. Blender, 3DS Max) this may be combined with telemetry information from cutting machines (distances, seam heights, cutting rates etc) and then add additional visualization and environmental affects in game engine such as Unity or Unreal for example adding particle systems for dust etc. There are also industry specific specialist simulation and training engines. If you are working with an equipment manufacturer, equipment models are usually converted from CAD data, then UV mapped, textured and keyframed either in traditional 3D tools or use programmed procedural animation in engine. Though you can purchase equipment model assets - features have to match depending on training fidelity. Equipment control is usually matched to actual physical controllers - so you read the input from a real world controller and trigger the corresponding virtual model animation in the virtual environment. The mining equipment manufacturers have training teams who create the assets and could configure layouts for specific client scenarios. As KineticLensmen pointed out - its the last stage of customization for specifics that usually takes longest since its crucial for the spatial awareness of teams being trained.

by vnxli 6y ago

KineticLensman 6y ago

A company I used to work for built a large sim for an Australian mining company, using a commercial game engine and a large 'igloo' type structure as the interaction environment. This was chosen over VR because it was believed to be easier to have groups of trainees in the same environment. The trainees were equipped with physical 'torches' that were geometrically paired with the virtual environment so you could actually shine a pointer on objects. The main training tasks were spatial orientation in the large 3D maze that was the mine, as well as hazard recognition and safety /;emergency training. E.g. knowing which way to go if an explosion occurred and how to avoid being crushed by the large vehicles that moved through the mine. The main unexpected development cost turned out to be the high degree of customisation required of the virtual assets, in particular the signage. Every tunnel and shaft had individual location and safety info on signs, and these had to be correct or the training value was lost.

vnxli 6y ago
The high-fidelity replication is much easier for things like industrial control systems (ICS)because all that needs to be replicated is the operator interfaces. Those interfaces are usually intentionally bland and typically dated - meaning really simple, standardized graphical assets I can imagine a complete mine recreation would be a huge development effort - especially with how huge they get. Is that something that like a camera drone could scan? i guess it could get the topology figured out with some special software, but could it get the signs and landmarks too? Can drones fly effectively in a mine?
- KineticLensman 6y ago
  > but could it get the signs and landmarks too? IIRC the customers supplied the primary graphics content for the signs digitally from their own files; the underestimated development cost was finding and editing the appropriate instances of the signpost objects. The corridor rock walls were a generic texture. There were also lots of physical 'emergency exit this way' indicators (e.g. arrow shapes that could be felt by touch in darkness) on railings. These were instanced as well, but the direction had to be checked. > Can drones fly effectively in a mine? No idea but think lots of corridors with short Line Of Sight. Probably easier now to use Lidar or similar. [Edit] Also, some of the signs were dynamic, e.g. "Checked by X on date Y" and needed to be updated in a scenario-specific way.
- kortex 6y ago
  > Those interfaces are usually intentionally bland and typically dated - meaning really simple, standardized graphical assets This sounds kind of like "worse is better". it got me thinking about user interactions, both gui and physical objects, from the angle of "what if I had to make these as 3D assets?" Screens suffer from limited real estate. But what could eg a CI pipeline dash look like in VR? You could have a whole virtual "control board" like a power plant.
- ptrott2017 6y ago
  Re: Drones - not in the sense I think you are imagining, but you can pull data from some of the mining equipment and I have seen 3D scanners mounted in situ at test sites to capture specifics. (it didnt work well - got remodelled later). The general workflow I have seen would be familiar to anyone in game or animation pipelines. A team models from the mine layout map in traditional 3D tools (e.g. Blender, 3DS Max) this may be combined with telemetry information from cutting machines (distances, seam heights, cutting rates etc) and then add additional visualization and environmental affects in game engine such as Unity or Unreal for example adding particle systems for dust etc. There are also industry specific specialist simulation and training engines. If you are working with an equipment manufacturer, equipment models are usually converted from CAD data, then UV mapped, textured and keyframed either in traditional 3D tools or use programmed procedural animation in engine. Though you can purchase equipment model assets - features have to match depending on training fidelity. Equipment control is usually matched to actual physical controllers - so you read the input from a real world controller and trigger the corresponding virtual model animation in the virtual environment. The mining equipment manufacturers have training teams who create the assets and could configure layouts for specific client scenarios. As KineticLensmen pointed out - its the last stage of customization for specifics that usually takes longest since its crucial for the spatial awareness of teams being trained.