I resent: the difference between 3D and virtual reality

Training drivers at the Chinese manufacturer Sapsanov. They took the head car with the driver's cab, copied all the devices and added a “view to the windows” using 3D screens.

I do virtual reality technology for engineers and for staff training. These are such systems where you can personally walk along an oil platform or a nuclear power plant, work out measures in the event of an accident in practice, and tighten the Main Valve with your working hands in gloves.

So, customers regularly confuse terminology and technology, in which they are very helpful, let's say, not quite professional market players. I would like to clarify and once again sort through what is what. I’ll say right away that after the first trial of the immersion system, all questions disappear, but here I’m not even able to convey feelings, so I’ll write words.
')

Myth 1: 3D pictures are not virtual reality.

Regular 3D pictures on a computer are far from being a virtual reality. Yes, there are models, render and everything, but the question is how it is perceived. Training in such a system of a nuclear power plant employee is not very far from the speed and usefulness of training on posters. The fact is, again, that there is no immersive effect, for the sake of which the whole forest with BP is fenced.

At sites of increased responsibility, it is assumed that within the framework of emergency situations on a simulator, you get a more or less accurate idea, not by logic, but “on the skin” as a whole. You know, it’s just like the pilots of an airplane - at first the transition from a computer simulator to a simulator with realistic controls and tilt modeling drastically upsets all the skills. It seems to fly well through the pictures, and here it seems for the first time. And for the second time, a similar jump occurs when moving from a simulator to a real aircraft.

You can learn from the pictures, but it is much harder and longer, plus it does not give the desired effect. Why? Because you will not be immersed in what is happening. When I ran along with everyone from the oil platform where the explosion occurred, I visually remembered the road, turned the valves with my hands, absolutely knew my position in space and the dimensions of various things, clearly saw everything in the volume. And in 3D on the screen it is not even possible to estimate the distance to something with eyes, not to mention other things. And this in the simulation is often very, very important. In MEI, for example, students put their heads in a turbine that “works” in terms of CAD-models and assemble and disassemble everything.


By presentation with such pictures, of course, you can guess how the train engine is arranged. But the sensations are compared to what you would actually see in real life and disassemble yourself with your hands, like heaven and earth. Students MIIT Railways work here with such models and play in the “simulator of the first person” almost every day.


From our data center

3D movie is not BP

“Ok,” say customers who understand this moment. - Let's make a 3D movie. We saw 3D movies in cinemas, very impressive. It is necessary to do a briefing on the evacuation or measures in case of an accident is the same. It will turn out cool, we even see for ourselves. "

The problem is that film and controlled reality are two different things. For example, in the second case there are scenarios that can trigger with different probabilities, or the trainer can trigger various developments. In the immersion system of BP, you personally do everything that is needed for, for example, evacuation. Run in the right direction, work with the necessary devices and tools, in the end you get instant feedback when making mistakes. It's like a game that you want to pass, but in which there is freedom of action. Naturally, games teach much better than movies.

Tests of our Western colleagues have shown that in the film, the sequence of actions is memorized very poorly.

The film is cool, but for real learning, we need systems where a person does everything himself. I do not know a single pilot, who learned to fly on the series.

Stereo plus mouse and keyboard - not yet BP

The third problem is that at the moment of understanding why virtual reality is still needed, the customer decides to stay on the stereo system with conventional controls. For example, a mouse and keyboard. Feeling, of course, is already better. From experience, I will say that, for example, our 3D model of a data center is very good to run through it in Counter-Strike. Of course, we began to orient ourselves in all its corners with our eyes closed, but this still does not help much during the development of actions for potential emergency situations. Because you have to go with your feet to the data center and correct the situation with your own hands already in place.

The keyboard and mouse is a barrier that prevents you from moving from the simulation mode to the mode when, standing upright, walk naturally along the object and remember all the actions kinesthetically, not visually. That is, the transition from visual memory to mechanical, motor skills, if you like. And the last - exactly what is needed for such training in emergency situations. What is missing in motor skills will be immediately forgotten at the first siren sounds. Or incorrectly done. Or at the wrong time. Or a person will painfully think before each step, translating logical experience into practical movements.

Sitting at a computer and typing with a mouse does not give a complete impression. When you run through an oil object in a virtual environment and something happens to you - a sound is heard, you can feel the valve. Instead of mechanical memory (where to click) there is a memory of what and how to do, at what distance in real scale which object is located from another.

World practice has shown that working out scenarios in a virtual reality environment is one of the best ways to transfer critical knowledge from the older generation to the younger. An old experienced nuclear scientist comes with a young one to an object and shows what is what. And then it launches the script of one of the accidents and looks at what it says. And I must say that young people are especially well “saved”. Learning is fast, and the process of knowledge transfer becomes more fun and efficient.

Occulus and similar systems are not industrial solutions.

“Ok,” says the customer. - It is clear, I walked along your cube, unscrewed some figovin from the turbine, put it in my pocket. But when I got out of virtual reality, something was gone. All clear. Let's do it with us, only on BP helmets - I recently wore this at a shopping center. The very thing.

The problem is that virtual reality helmets are such a thing, from which in 10–15 minutes it will be impossible to make you sick. Plus, even in the most modern helmets, large beautiful pixels are still visible, which do not give a normal focus on what you need.
One more thing. In the narrow viewing angle, the cerebellum feels that something is wrong. It's like in a car to play on the phone in "Karmageddon": like the movement of the car in reality and the control of your car in the game are not connected, but no, the motor skills suffer. And for a long time you can not play normally.

Helmets are good for the consumer segment. But if you drive for hours of training (and the elementary evacuation is worked out 6 hours to complete automatism), people will simply go crazy. Prepare paper bags.

Here are typical pros and cons of virtual reality rooms and head mounted displays:

• Rooms are more expensive, helmets are much cheaper.
• For rooms you need a special room, for helmets - no.
• A room is harder to transport from place to place, a helmet is easier.
• Helmets give low resolution, rooms high.
• The rooms have a joint work on the object (the student and the tutor in the same physical room and the tutor can almost lead the student’s hand). There is no such thing in helmets.
• The rooms have the ability to move freely, which dramatically increases the useful mechanical memory. In helmets - just turn your head.
• An open environment is modeled in rooms, and helmets are always a tunnel of vision.
• Rooms are equipped with accurate sensors for positioning objects inside, helmets often rely on accelerometers with high errors. Hence the difference in interactivity and accuracy of actions.
• Helmets give the feeling of dizziness and confined space, the room - no.
• Current helmets are severely limited in functionality and performance, the bottleneck of the rooms is the controller (laptop or cluster), which allows them to be used for years for different projects.
• Rooms occupy much more space during storage, helmets are easily removed to the warehouse.

What is virtual reality?

BP is a coherently working set of content systems, a projector, glasses, a synchronizer for flickering points and a controller (powerful computer or cluster). Properly assembled BP system allows you to get at the facilities of increased responsibility the main thing - to teach staff to make decisions instantly in case of an emergency. At a number of industrial facilities, a difference of 3-5 seconds can be decisive and cost not even a couple of million dollars (the cost of equipment), but dozens of human lives. That is why everything that allows you to transfer the experience of an emergency as accurately as possible deserves attention.

Of course, if there is an opportunity to work out “in kind” of an emergency, this should be used. But the only major stand of this kind known to me is a copy of the ISS (it was previously a copy of MIR), where future crews are undergoing training. And if there is depressurization somewhere, the guys won't be able to breathe either - such a BP is not yet able. But, of course, where it is impossible to pick up and copy for training purposes of an NPP, an oil platform, any industrial object (for example, a workshop for building airplanes or cruisers, mining production or something else), the BP technique is used as the closest one. Plus "physical" simulators for individual nodes.