4 ms·

> I don't. I believe in AR but not short-term. I think AR will go into a "winter" soon and will be picked up by the mainstream some years later after key techno

> I don't. I believe in AR but not short-term. I think AR will go into a "winter" soon and will be picked up by the mainstream some years later after key technology challenges are solved. I think AR (glasses) is at where VR was in the nineties. > Currently AR has two form factors, phone and glasses. > Phone-based AR is gimmicky. It can be fun and can work in some scenarios but it's monoscopic, has a ridiculous field of view, requires holding your phone in the air, and it's not really augmenting reality anyway, it's augmenting a video stream of reality (increased latency, reduced dynamic range, motion blur, etc.). > Glasses-based AR will be great but currently has fundamental challenges with field of view, latency, optics, and integration of augmentations in the real world (e.g: procedural objects dynamically matching physical obstacles). It requires new inventions in hardware and software to fully realize the vision we have of it. > On the other hand VR has a clear technological path for the next decade into human-like field of view and visual fidelity. The full enclosure means we can get +200° and fit eye tracking cameras in the headset. Foveated rendering, only render what you actually look at at full res. The advent of ray tracing hardware (easier foveated rendering and non-linear optics). There is no fundamental barrier to get from here to 60 pixels per degree at 200° fov. I 99.9% agree with this assesment, particularly with respect to the timeline of product maturity/adoption. However, I also agree with Andybak above, that AR & VR applications are complimentary, but largely different. One will not displace the other, but the technologies will merge such that you can seamlessly transition from AR to VR nd back using the same hardware (likely Virtual Retinal Display-based displays) and operating system. I use VR in my job (evaluating power plant designs), and for that, I need to be immersed in a 3D model as opposed to AR where not being fully immersed would be unnecessarily distracting. On the other hand, after a plant is constructed, AR will be extremely useful (not yet, but eventually) in identifying where the actual design departs from the contractual model, or for training (new people learning how to operate the plant), or for monitoring (seeing overlays of temperature/pressure/planned maintenance data/etc., and so forth). VR right now is mature in that the quality of display and ease of UI make it useful. It's still a pain to set up and get the models imported. It also still suffers from relatively low resolution. Resolution is still too low to say, emulate a computer monitor (i.e. present a virtual desktop computer monitor and use it at normal seated viewing distance - which requires a minimum 35PPD). For my needs, it's hard to read even large text on virtual signage in VR (you have to get VERY close to the virtual sign to read it). So, unlike the above comment, I think the VR industry has quite a bit of work to get from it's current sub-35PPD to a 60PPD display approaching the limits of human eye resolution.

7952 8y ago
My company is trying to push use of VR in engineering, and I have been a bit confused by the desire. We have had 3D models for years that could be viewed on a normal desktop, and peoe usually prefer 2d plans. The improved UI seems to be very attractive to people though. I wonder if a major benefit of VR is that it forces us to create visualisations that are responsive. It is easy to get away with slow rendering on a desktop, but you can't on a headset. The software has to be realtime, and the model optimised for that.
- mojomark 8y ago
  For me, we are concerned with designing our plants to ensure humans can interact efficiently & safely with the systems (power plants) we're designing. Here are some benefits for me of VR over 2D viewing of 3D models: 1. Visibility assessments. In VR you can get a true sense of Field of View (FOV). For example, operators want maximum visibility through windows in control stations (or in a flight control tower, or sitting in the driver seat of a car). In 2D viewing, you can adjust the perspective settings to be anything and you simply can't match a humans natural FOV, which in addition to matching true perspective, you also capture the real time dynamics of how visibility changes as a user's body/head move. I've seen control stations completely redesigned after being built because the customer didn't like the experience (a risk largely mitigated with VR). You can't put 'control stations' in glass bubbles. It just isn't feasible. You need reasonably-sized windows and mullions, and there are many obstructions to contend with. 2D visualizations of 3D walk-through models simply don't accurately capture what's a VR experience can. We've actually found that, concerns about visibility identified from 2D visualizations are often discounted when we check it out in VR, just because the way and how quickly people move around naturally and subconsciously improves visibility. 2.) Ergonomics (Human-Machine Interface). In VR, you can literally reach out with your physical hands and determine if a valve, or a button is safely reachable by an operator. In initial design, you rely on standards/rules of thumb for placement of these items. However, as the design matures, you inevitably end up with pipes/wires/beams that complicate matters - it's impossible to right rules for every possible scenario. Historically, the only way to be sure. 3. Efficiency. I've found that user's can intuitively move around in a VR model much faster than they can with a screen/mouse. Especially, when you have a complex nest of pipes you're trying to evaluate. Using mouse/joystick/keyboard clicks and rotating takes people much longer than simply bending down and turning their head to see an object hiding behind a pipe or other object. 4. Remote Inspections. Once the plant (or whatever) is built, it's typically not near the office of the engineering firm that designed it or customer's headquarters that paid for it. Historically, when you get reports of people complaining about a design, you go into a 2D model first to check it out. If you can't get a good sense of the issue (typically because you can't get a true feel for the ergonomics/visibility aspects of the issue), you buy a plane ticket, rental car, hotel, and go an inspect in the plant in person. However, VR is good as assessing ergonomics and visibility issues (as described above), so you can use VR to avoid the need for travel. This has occurred already on a few occasions. We currently only use static models, but smarter people can begin encoding dynamic features (e.g. opening doors, removing panels, etc.). I'm sure other people are doing this already. In conjunction with the ergonomics/visibility aspects above, I expect this will greatly enhance VR benefits.

7952 8y ago

My company is trying to push use of VR in engineering, and I have been a bit confused by the desire. We have had 3D models for years that could be viewed on a normal desktop, and peoe usually prefer 2d plans. The improved UI seems to be very attractive to people though. I wonder if a major benefit of VR is that it forces us to create visualisations that are responsive. It is easy to get away with slow rendering on a desktop, but you can't on a headset. The software has to be realtime, and the model optimised for that.

mojomark 8y ago
For me, we are concerned with designing our plants to ensure humans can interact efficiently & safely with the systems (power plants) we're designing. Here are some benefits for me of VR over 2D viewing of 3D models: 1. Visibility assessments. In VR you can get a true sense of Field of View (FOV). For example, operators want maximum visibility through windows in control stations (or in a flight control tower, or sitting in the driver seat of a car). In 2D viewing, you can adjust the perspective settings to be anything and you simply can't match a humans natural FOV, which in addition to matching true perspective, you also capture the real time dynamics of how visibility changes as a user's body/head move. I've seen control stations completely redesigned after being built because the customer didn't like the experience (a risk largely mitigated with VR). You can't put 'control stations' in glass bubbles. It just isn't feasible. You need reasonably-sized windows and mullions, and there are many obstructions to contend with. 2D visualizations of 3D walk-through models simply don't accurately capture what's a VR experience can. We've actually found that, concerns about visibility identified from 2D visualizations are often discounted when we check it out in VR, just because the way and how quickly people move around naturally and subconsciously improves visibility. 2.) Ergonomics (Human-Machine Interface). In VR, you can literally reach out with your physical hands and determine if a valve, or a button is safely reachable by an operator. In initial design, you rely on standards/rules of thumb for placement of these items. However, as the design matures, you inevitably end up with pipes/wires/beams that complicate matters - it's impossible to right rules for every possible scenario. Historically, the only way to be sure. 3. Efficiency. I've found that user's can intuitively move around in a VR model much faster than they can with a screen/mouse. Especially, when you have a complex nest of pipes you're trying to evaluate. Using mouse/joystick/keyboard clicks and rotating takes people much longer than simply bending down and turning their head to see an object hiding behind a pipe or other object. 4. Remote Inspections. Once the plant (or whatever) is built, it's typically not near the office of the engineering firm that designed it or customer's headquarters that paid for it. Historically, when you get reports of people complaining about a design, you go into a 2D model first to check it out. If you can't get a good sense of the issue (typically because you can't get a true feel for the ergonomics/visibility aspects of the issue), you buy a plane ticket, rental car, hotel, and go an inspect in the plant in person. However, VR is good as assessing ergonomics and visibility issues (as described above), so you can use VR to avoid the need for travel. This has occurred already on a few occasions. We currently only use static models, but smarter people can begin encoding dynamic features (e.g. opening doors, removing panels, etc.). I'm sure other people are doing this already. In conjunction with the ergonomics/visibility aspects above, I expect this will greatly enhance VR benefits.