1
$\begingroup$

My VR machine isn't the plug-it-to-your-brain type, but instead it's more like a "reality sphere" around your body that have some kind of gravity machine to keep you in the center of the sphere and some kind of gray goo that can reshape to create physical contact experience (imagine the robot from Terminator series). It's designed for entertainment and education purpose and aimed for general public.

As you know it, the worst case scenario of VR is to be stuck inside of it. So I make a fail-safe to prevent this from happening.

To quit the simulation you can either: push the button on your VR watch, OR say 'WOLUBA', OR simply close your eyes for 10 seconds.

The machine also detect your vital sign and will automatically stopped itself and contact the authority should you faint... Or die. It's also create a barrier around your body to reduce the force you might get to a non-harmful level and prevent the gray goo from entering your body.

What do you think of this fail-safe design? Do you find any flaw in it? Do you thing there are other thing that I forget to consider? Do you have a better fail-safe design?

$\endgroup$
  • 1
    $\begingroup$ Do you have a particular reason to use a free-floating goo, as opposed to a more traditional rig like a skintight suit? The goo seems to be the main problem from a safety and usability perspective (and it sounds like a pain to wash off afterwards). $\endgroup$ – Cadence Nov 28 '19 at 0:00
  • $\begingroup$ Have someone to supervise or go live streaming at least there is a high chance to get swatted!🚔 $\endgroup$ – user6760 Nov 28 '19 at 0:23
  • $\begingroup$ @Cadence It's because the goo can shape to any physical thing. For example, if you create a simulation where you walk through a stairs, the goo will shape itself to mimic stairs while the gravity machine kept you from actually walking through stairs so you won't get out from the "reality sphere". Imagine the treadmill; you're running, but going nowhere. I don't think the goo will leave a stain in your clothes since they are a smart nanomachines. BTW I don't find a correlation between VR and skintight suit. Could you give me a referrence? $\endgroup$ – arlilo Nov 28 '19 at 0:45
  • 3
    $\begingroup$ True, you can use the gel to create "props", but do you need to? That is, if the player can't see them, all they have to go on is the impression on their skin - the object doesn't need to have any more reality than that. As for VR suits - there have been real experiments in that direction, but I think the best recent sci-fi reference I can offer is the novel Ready Player One (unless I've badly mis-remembered it). In that book it was an optional extra though, a sort of peripheral to the VR setup. $\endgroup$ – Cadence Nov 28 '19 at 0:59
  • 1
    $\begingroup$ The series Otherland by Tad Williams included hand gestures that involved spreading the fingers painfully wide to trigger functionality that one does not want to cause by accident. $\endgroup$ – Cort Ammon Nov 28 '19 at 5:23
5
$\begingroup$

Supplemental Reading

Yes, I'm starting my answer with the supplemental reading, because in this case, the supplemental reading is amazing.

From a worldbuilding perspective, what you are designing is very similar to a Star Trek holodeck (or holosuite). Many of the stories focusing on the holodecks on Star Trek: TNG, Voyager, and the holosuites on DS9 are excellent sources of how their "safeties" work, how they malfunction, what the failure modes are, and why someone might want to override them (and how to limit access to that feature).

The Trekkie's trove of information, MemoryAlpha, has a page dedicated to this: Holodeck safety protocol

Specific Failure Modes

I'll go through each of your proposed safety measures and explore how they might fail:

Push the button on your VR watch

Electronics fail, mechanical buttons fail, power supplies/batteries fail, the signal from the watch might not reach the receiver. Maybe you fall in VR and break your wrist, and the watch along with it, and now you can't use the watch to get out and seek medical attention.

Say 'WOLUBA'

In Star Trek, this would be "Computer, Exit" or "Computer, Arch". Those voice commands failed frequently, for a number of reasons, but usually due to some malfunction in the computer, or a rogue AI taking control. It's important to note that when this type of failsafe fails, other related systems may fail as well, so it's entirely possible you find yourself trapped in VR with the other physical safety limitations malfunctioning as well.

OR simply close your eyes for 10 seconds.

Assuming your technology has absolutely perfect eye tracking and would detect this condition 100% accurately, my guess would be that the failure mode here would be users disabling that feature entirely, because it would be a pain in the butt. Imagine, say, enjoying a simulation of Beethoven performing his 9th Symphony only to close your eyes to feel the music... and... whoops. And with the feature disabled, users have one less failsafe keeping them from danger.

Vital signs

This one is good, if it is accurate, and adjusted for the baseline of the individual. It would need access to one's medical history, so as not to trip due to, say, tachycardia in someone who has a naturally high heart rate, or high blood pressure in someone with hypertension. Limits would either have to be very permissive, and trigger only on serious conditions, or users would have to deal with it triggering for benign "nuisance" reasons, such as momentary blackouts one might experience doing high-g flight training.

Privacy also needs to be considered, with medical history, notifications, and how those are transmitted.

Barrier around your body

Aside from the local gravity, this barrier might be the most difficult thing to pull off that you have suggested. Given the grey goo is already under perfect control, you might be better off putting all of the smarts into that. You can even protect against "passive" injury, such as the person tripping over their own feet and smashing their forehead into the corner of a table, by detecting the motion of the user and "softening" any objects they are about to collide with. This, to me, seems easier and more direct than introducing an entirely new force field system with entirely new failure modes. Especially since a typical sci-fi force field won't protect you from concussive brain injury in the above table scenario.

You can prevent the grey goo from entering someone's body by having it simply move itself away from/out of any orifices it may find itself in (or any new orifices it may be tempted to create). With the proposed level of tech, it would be trivial (and necessary) to 3D scan the body of the user in space anyway, so at all times the grey goo would know if a given coordinate in space is "inside" or "outside" the user.

Better Ideas?

You asked for better ideas. I think your ideas are generally fine, but what they lack is a hard, physical interlock. For example, in real life VR, you have the absolute ability to end the simulation by simply removing the headset. The software and hardware could be completely malfunctioning, yet that always works.

Your simulation is far more immersive, however, so any sort of physical interlock would take away from the simulation, and may be undesirable. Your VR watch is not an example of this, because it is still a "soft" exit, as it only sends a signal and relies on the software/hardware to do the right thing.

One way around this is to put someone else in charge; an observer on the outside that is responsible for stopping the sim on your command (or if certain parameters are exceeded). For a system this complex and large, it's likely it would be a staffed installation anyway. Obviously this person can fail as well, but so can you, if you are unconscious or otherwise incapacitated, so I judge it to be about the same risk. You probably need both exit methods. For example, you could have the location of the exit door always visible, like the "Exit" signs legally required in public buildings around the world.

It also couldn't hurt to set an absolute time limit for the simulation so the user doesn't die of thirst if all of the other exit protocols fail. You can make this malfunction-proof by running the sim off the grid and on battery power, so once the batteries are dead, the simulation stops. I suspect the energy requirements for this would be well into the megawatt range, though, so it's up to you if your universe has batteries with sufficient energy density to power this simulation for the required time.

| improve this answer | |
$\endgroup$
  • $\begingroup$ The battery suggestion is excellent. $\endgroup$ – Joe Bloggs Nov 27 '19 at 23:56
  • 2
    $\begingroup$ Medical privacy is a huge concern. You don't want to be on the hook for making your software HIPAA-compliant (or future-HIPAA-compliant) if you can at all avoid it. It may be easier if you make a dedicated hardware module that reads in vital data and outputs only a "go/no-go" signal. $\endgroup$ – Cadence Nov 28 '19 at 0:23
  • 2
    $\begingroup$ As an alternative to running it off of massive batteries, you can run it off of mains controlled by a relay, which is run off a battery. When the battery runs out, the relay opens, and no more power. This isn't quite as safe, since the relay might fail closed, but if you use several of them, you can get a pretty good balance between safety and practicality. (freefall.purrsia.com/ff2300/fc02206.htm used this idea...) $\endgroup$ – Matthew Nov 28 '19 at 16:34
5
$\begingroup$

You'll also want to restrict the sharpness of the goo, limit contact between the goo and the face to prevent suffocation, and include some programming to prevent the goo from putting light pressure on certain combinations of pressure points to prevent harm to the body.

And you might want to limit sound and light intensity to safe levels, and include safeguards to prevent certain seizure-inducing activities (flashing lights, etc).

Also, typically, in addition to software safeguards like the ones you've mentioned, equipment like this IRL will include a physical safeguard, like a cord which hangs from the ceiling that you can pull in case of a software failure, or a deadman-switch that you have to hold firmly between your fingers or teeth, and if you ever relax your grip then the machine loses a hard-signal and turns off. Software just can't always be trusted.

| improve this answer | |
$\endgroup$
  • 3
    $\begingroup$ +1 for mentioning dead man switches especially. In engineering terms, this is the very definition of fail safe - that if the system loses contact, power, etc. then it fails in a way that is a safe manner for those around it. Ideally, the goo would lose cohesion and fall into grates on the bottom of the VR sphere at a loss of power or loss of signal from a dead man switch of some form. $\endgroup$ – Tim B II Nov 27 '19 at 22:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.