Help - Search - Members - Calendar
Full Version: Security Cameras and Perception Rolls
Dumpshock Forums > Discussion > Shadowrun
AndyZ
Let's say that I've got a security camera set up and staring out from my safehouse. Meanwhile, the evil NPC Slimyman is trying to sneak past the camera in order to evade notice and get into my place unopposed.

Slimyman will roll Agility+Infiltration in an opposed test. What will I roll? Obviously Perception, but the camera's rating? My Intuition?
Draco18s
QUOTE (AndyZ @ Mar 3 2010, 03:38 PM) *
What will I roll? Obviously Perception, but the camera's rating? My Intuition?


This is why the whole OR4 for cameras thing with regards to (physical) invisibility bugs me. A dude watching a camera somehow instantly knows the mage is there (perception threshold 1, regardless of what dice are being rolled) because the spell wasn't strong enough.

I think, however, that you're rolling Intuition + Camera Rating (Tool taking place of Skill, because you can't "look everywhere" only where the camera is aimed) or Perception + Intuition + Camera (the usual Skill + Stat + Tool).
D2F
QUOTE (AndyZ @ Mar 3 2010, 09:38 PM) *
Let's say that I've got a security camera set up and staring out from my safehouse. Meanwhile, the evil NPC Slimyman is trying to sneak past the camera in order to evade notice and get into my place unopposed.

Slimyman will roll Agility+Infiltration in an opposed test. What will I roll? Obviously Perception, but the camera's rating? My Intuition?


If it is just a regular video camera, you perception+intuition, adjusted by the applicable situational and visbility modifiers.

To give an excemple:

You have 4 cameras set up across your are (2 at the doors, 2 outer area in this excemple). You have the AR display for both up, while watching an action Trid during your after hours. You are not overly paranoid, as you have not been warned beforehand and no reason to expect any to sneak into your house. It is late evening and the light is dim. The street lights haven't turned on, yet.

The roll would be:

Perception + Intuition - 2(partial light) - 2 (distracted) - 2(interfering sight/sound(watching a trid)
In other words: Perception + Intuition -6
threshold would be either (2), if the target wasn't even trying to sneak, or their infiltration hits, if they are sneaking.
Ascalaphus
The problem is that you'd like to aggregate the cameras somehow. You don't make a roll to sneak past a single camera; you roll to sneak past all of the cameras. Meanwhile the cameras are set up in a pattern optimized to spot you (hopefully.)

So the check to spot someone sneaking by shouldn't be four separate checks, but a single check, improved for having multiple cameras.



SR really has a problem in getting the sensors to work together (see my comments about drone Sensors)
D2F
QUOTE (Ascalaphus @ Mar 3 2010, 11:35 PM) *
The problem is that you'd like to aggregate the cameras somehow. You don't make a roll to sneak past a single camera; you roll to sneak past all of the cameras. Meanwhile the cameras are set up in a pattern optimized to spot you (hopefully.)

So the check to spot someone sneaking by shouldn't be four separate checks, but a single check, improved for having multiple cameras.



SR really has a problem in getting the sensors to work together (see my comments about drone Sensors)


And this is where you're wrong (in my opinion). The cameras are not snetient. They are usually just fed through to the security station, where they can watch them on multiple AR overlays or on actual trideo screens. The problem is: the more cameras there are, the more distracting they are, the more likely it is for you to miss that one camera angle that shows the intruder.

Now, of course the higher up security goes, the more sophisticated systems become. The security system might be tied together with motion sensors, heat sensors, seismic sensors, ultrawideband radar, atmospheric sensors, MAD scanners, NLD scanners... you get the gist. The difference that then happens is that the various systems try to detect the intruder (which you could roll as a teamwork test, theoretically), and if successful would notify you. You could then check the specific camera in detail and use the appropriate perception modifiers (+3 for actively looking for it, +vision enhancements + 2 for object standing out, for a whopin total of +5 to +8 dice on your perception roll) against the appropriate threshold (2 or the infiltration hits).

ince the OP question was just about the camera, though, the more you have the less likely you are to see the intruder on camera. Just imagine the difference between having to watch 3 different screens and 20. And ask yourself whether you would constantly pay concentrated attention to all of them over the course of your 8 hour shift...
You might catch an intruder on the screen in the corner of your eye, bu that's what the perception modifiers are for in the first place.
Ascalaphus
But with more cameras, you can cover more angles - there are fewer places in the room where you can effectively hide. That should measure into the difficulty of spotting an infiltrator.


I'll agree that a clever system doesn't require one guard to watch all the screens all the time, and uses some software to check for untoward movements to focus the guard on. But especially in that situation will multiple cameras really be of value.
D2F
QUOTE (Ascalaphus @ Mar 4 2010, 02:50 PM) *
But with more cameras, you can cover more angles - there are fewer places in the room where you can effectively hide. That should measure into the difficulty of spotting an infiltrator.


What it does, is affect the Infiltratin roll, not the perception roll. For the perception roll, the more cameras you have the worse you are off (even though rules wise, even two cameras already impose the highest negative impact on your perception roll)

QUOTE (Ascalaphus @ Mar 4 2010, 02:50 PM) *
I'll agree that a clever system doesn't require one guard to watch all the screens all the time, and uses some software to check for untoward movements to focus the guard on. But especially in that situation will multiple cameras really be of value.


No, they won't. You seem to confuse "Sensors" and "Cameras". A camera doesn't detect anything. It simply records visual information. You still need an external source (be it a guard or image recognition software, or pattern recognition software or effing MS excel to analyse that visual information and derive a proper assessment from it. It is the quality and power of that additional ressource, that interpreting agent, that determines, whether or not you detect the intruder.

As an easy rule of thumb: Cameras don't set off alarms. Guards do. (note: guard in this aspect can be anything from a rent-a-cop to a highly sophisticated military grade pattern analysis supercluster)
FriendoftheDork
QUOTE (D2F @ Mar 3 2010, 09:45 PM) *
If it is just a regular video camera, you perception+intuition, adjusted by the applicable situational and visbility modifiers.

To give an excemple:

You have 4 cameras set up across your are (2 at the doors, 2 outer area in this excemple). You have the AR display for both up, while watching an action Trid during your after hours. You are not overly paranoid, as you have not been warned beforehand and no reason to expect any to sneak into your house. It is late evening and the light is dim. The street lights haven't turned on, yet.

The roll would be:

Perception + Intuition - 2(partial light) - 2 (distracted) - 2(interfering sight/sound(watching a trid)
In other words: Perception + Intuition -6
threshold would be either (2), if the target wasn't even trying to sneak, or their infiltration hits, if they are sneaking.


As someone who spends a inordinate amount of time gazing at security cameras (sometimes watching flicks on a different screen)...

AT the first part of your post I agree with it. Since the guard is not paying complete attention to the camera the "distracted" and watching a trid, a perception check is surely in order, with intuition. The check is based on an attempt to actually notice movement on the screen/image, not so much if the intruder is visible on the camera feed at any one time. Thus sensor rating or camera rating is fairly irrelevant.

But the added penatlies and threshold makes this an impossible task. There is no way a normal person has a chance of succeeding at all, and even skilled security guards would fail. A -6 perception test is more or less the equivalent to shooting in pitch dark...and a threshold of two basically means you need least 12 dice in order to have a decent chance at success - 12 dice is pretty much the pinnacle in unaided perception tests.

I'd change the roll to perception+intuition -2 for distraction, -2 for interfering sight/sound. -2 for poor light is valid, but who has set up cameras lacking night vision when monitoring them in the late evening night?Also, I might even be inclined to either remove the interefering sight OR the distraction penalty if the person felt somewhat obliged to glance at the screens from time to time... or combine the penalties to become -3. Threshold would be 0 for fairly obvious movement and 1 for either slow movement or very quick movement (lasting less than a second).

Also, there might be a case for the person to stand out in some way, although not likely for a stealthy intruder. In any case such a ninja would probably either try to avoid the cameras line of sight completely (if possible) or disable them somehow. But you can't really sneak past a camera if someone is actually watching it. Distance might also factor in. The first 10-20 meters wouldn't matter much, but if you had the camera set up along a street far away peope sneaking across it would be very hard to detect.

In any case sneaking past cameras while someone is watching is more about luck than actual sneaking skills -when sneaking you can usually see the people you are trying to evade, anticipiate their head movement etc. but with a camera that is impossible unless it is the silly Deus Ex/Metal gear rotating kind...
D2F
QUOTE (FriendoftheDork @ Mar 4 2010, 04:48 PM) *
But the added penatlies and threshold makes this an impossible task. There is no way a normal person has a chance of succeeding at all, and even skilled security guards would fail. A -6 perception test is more or less the equivalent to shooting in pitch dark...and a threshold of two basically means you need least 12 dice in order to have a decent chance at success - 12 dice is pretty much the pinnacle in unaided perception tests.


It was a specific excemple. And it was pretty much a worst case scenario.

QUOTE (FriendoftheDork @ Mar 4 2010, 04:48 PM) *
I'd change the roll to perception+intuition -2 for distraction, -2 for interfering sight/sound. -2 for poor light is valid, but who has set up cameras lacking night vision when monitoring them in the late evening night?Also, I might even be inclined to either remove the interefering sight OR the distraction penalty if the person felt somewhat obliged to glance at the screens from time to time... or combine the penalties to become -3. Threshold would be 0 for fairly obvious movement and 1 for either slow movement or very quick movement (lasting less than a second).


-4 would the usual modifier in this case. Keep in mind, though, that gunfire and similar obvious factors modify that roll further:

Gunfire and similarly obvious facors reduce the nescessary threshold to 1
Object standing out (like a dark shade on the otherwise light background of the company parking lot) adds a further +2 to the dice pool (reducing the overall modifier to -2).

The camer could laso have additional Vision Enhancemen, which would add the enhancement rating to the dicepool (changing the final modifier to anywhere from -1 to +2)

At this point it should also become obvious how important other vision enhancements become, like Thermographic vision, urltrasound sensors or ultrawideband radar.

The difference beteen a -6 dicepool on a 2 threshold and a +2 diece pool and a 1 Threshold should be obvious.

QUOTE (FriendoftheDork @ Mar 4 2010, 04:48 PM) *
Also, there might be a case for the person to stand out in some way, although not likely for a stealthy intruder. In any case such a ninja would probably either try to avoid the cameras line of sight completely (if possible) or disable them somehow. But you can't really sneak past a camera if someone is actually watching it. Distance might also factor in. The first 10-20 meters wouldn't matter much, but if you had the camera set up along a street far away peope sneaking across it would be very hard to detect.


That is what the infiltration skill is for and why it modifies the perception threshold. Infiltration is not a meas to become magically invisible, but to know how to avoid cameras, limit your own emossions (including sound), knowing how to keep your profile against a dark background and so forth.

Also, distance has its own dice pool modifier for the perception roll:

target far away: another -3 dice pool modifier

QUOTE (FriendoftheDork @ Mar 4 2010, 04:48 PM) *
In any case sneaking past cameras while someone is watching is more about luck than actual sneaking skills -when sneaking you can usually see the people you are trying to evade, anticipiate their head movement etc. but with a camera that is impossible unless it is the silly Deus Ex/Metal gear rotating kind...


I disagree. Sneaking past cameras is not about luck, but how to avoid showing up on them in the first place.
Ascalaphus
QUOTE (D2F @ Mar 4 2010, 03:12 PM) *
What it does, is affect the Infiltratin roll, not the perception roll. For the perception roll, the more cameras you have the worse you are off (even though rules wise, even two cameras already impose the highest negative impact on your perception roll)


In another thread we went round and round about Infiltration modifiers. Turns out there aren't any; there are only Perception modifiers. Because otherwise you'll have to start rolling separate Infiltration checks against all possible kinds of observers.


QUOTE (D2F @ Mar 4 2010, 03:12 PM) *
No, they won't. You seem to confuse "Sensors" and "Cameras". A camera doesn't detect anything. It simply records visual information. You still need an external source (be it a guard or image recognition software, or pattern recognition software or effing MS excel to analyse that visual information and derive a proper assessment from it. It is the quality and power of that additional ressource, that interpreting agent, that determines, whether or not you detect the intruder.

As an easy rule of thumb: Cameras don't set off alarms. Guards do. (note: guard in this aspect can be anything from a rent-a-cop to a highly sophisticated military grade pattern analysis supercluster)


A camera has a field of vision; Infiltration is mostly a matter of staying out of that field of vision. More cameras allows you to put more of the room in-vision. Therefore more cameras are helpful.

Even when you don't have any fancy software, multiple screens shouldn't present too much of a penalty; if you're watching at all, your attention will be drawn to the screens that display anything out of the ordinary.


I'm advocating the use of a set of cameras as a whole; the guard rolls to see if he spots the Infiltration on any of the relevant screens, not one roll per screen. All in all, multiple well-placed cameras should make the perception easier because more of the room is visible.

Of course, if you're serious about security, you combine cameras with motion sensors. If the motion sensor picks something up, the guard pays especial (= not distracted) attention to those camera feeds.

(Also, eliminating false positives from vermin is a realistic possibility with the use of drones, special animals and heck, low-Force spirits on extermination duty.)
Udoshi
Remember that, while using electronic sensor devices, you may substitute Sensor for Intuition.

So a camera network would generally roll 3 to 9 dice to spot things. Rating 3 standard, rating 6 is cheap, and vision enhancement 3 is cheaper. Since this is SR, and AR is very prevalent, it would most likely flag the guard when it spots people in areas they shouldn't be in.

This is before you set it up with an Agent or Pilot program, networked and watching the building security. Or load the entire network up with Sensor softs. Because, yeah, people do that in SR. Security spiders can, and do, set up whole buildings to be riggable and remote controllable for just that purpose.
D2F
QUOTE (Ascalaphus @ Mar 4 2010, 05:11 PM) *
In another thread we went round and round about Infiltration modifiers. Turns out there aren't any; there are only Perception modifiers. Because otherwise you'll have to start rolling separate Infiltration checks against all possible kinds of observers.


My apologies. I still had our house rules in mind. Still, the infiltration skill covers the character's ability to circumvent detection by any observer, including electronic devices. And having more cameras doesn't make it more likely to detect a given intruder, it just gives you a chance to detect him in more areas, allowing you a perception roll in the first place. All of that is irrelevant to the thread, though.

QUOTE (Ascalaphus @ Mar 4 2010, 05:11 PM) *
Even when you don't have any fancy software, multiple screens shouldn't present too much of a penalty; if you're watching at all, your attention will be drawn to the screens that display anything out of the ordinary.


While that might be the case if there was zero movement anywhere (including your Trid), the information overflow hypothesis tears a huge chunk out of your assumption. The more information the human brain is presented with, the more likely it is to dismiss information.
The low rate of altruistic behaviour in urban environments is testament to that fact. This is an established psychological mechanism which specifically pertains to perception and as such is relevant in this context.


QUOTE (Ascalaphus @ Mar 4 2010, 05:11 PM) *
I'm advocating the use of a set of cameras as a whole; the guard rolls to see if he spots the Infiltration on any of the relevant screens, not one roll per screen. All in all, multiple well-placed cameras should make the perception easier because more of the room is visible.


Not necessarily. See above. Rather than increasing the number of cameras it makes much more sense to add additional functionality, like vision enhancement, MAD scanners, atmospheric scanners or Ultrasound, to improve the amount of information per screen (note that improving in this case does not mean simply adding, as that would be detrimental, but presenting the visual information in a different manner. Excemples include highlighting detected individuals ina bright green on an otherwise monochrome Thermal image, or providing auido clues along with flashing indicators on screen, or moving the AR segment of the camera in front of the user, effectively blockng the Trid he is watching).
All of those features would drastically change the perception roll and for good reason. Detecting a dark shape on a monochrome Image is unlikely. Detecting a flashing green shade on an otherwise gray image is much, MUCH more likely, especially when combined with the ride of the valkyries thundering from internal speakers.

QUOTE (Ascalaphus @ Mar 4 2010, 05:11 PM) *
(Also, eliminating false positives from vermin is a realistic possibility with the use of drones, special animals and heck, low-Force spirits on extermination duty.)


False positives are less about actual vermin, but about technical issues. The more tech you employ, the more lilkely it is that one of them glitches. That includes analytical software. It also includes the huamn brain.
D2F
QUOTE (Udoshi @ Mar 4 2010, 06:06 PM) *
Remember that, while using electronic sensor devices, you may substitute Sensor for Intuition.

So a camera network would generally roll 3 to 9 dice to spot things. Rating 3 standard, rating 6 is cheap, and vision enhancement 3 is cheaper. Since this is SR, and AR is very prevalent, it would most likely flag the guard when it spots people in areas they shouldn't be in.

This is before you set it up with an Agent or Pilot program, networked and watching the building security. Or load the entire network up with Sensor softs. Because, yeah, people do that in SR. Security spiders can, and do, set up whole buildings to be riggable and remote controllable for just that purpose.


A high end security surveillance system could look like this:

Mounted Camera (Rating 6): Vision Enhancement 3, Thermographic Vision, Ultrasound, Image link, Radio Signal Scanner (6), MAD Scanner (3), Motion Sensor, Olfactory Scanner (6) (Chemsniffer)
Sensor Software package: Facial Recognition Software (6)(checking recorded Faces with the databank of those allowed to be there), Visual Spotter Software (6) (tracking humanoid shapes, to highlight them), Wildlife Spotter (6) (to avoid false alarms from urban or local fauna)

That is one small, neat, mounted package, with a Signal Rating of 4, tied into the security system (including the ECCM 4).
Total Price: 17,850
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Dumpshock Forums © 2001-2012