BART killing exposes security gap – many train cameras are decoys | Hacker News Reader

BART killing exposes security gap – many train cameras are decoys

sfgate.com

81 pointsjohnhenry59 comments

Another problem with "dummy cameras" for PUBLIC institutions is that they are basically lying to the taxpayer about what level of security is being funded and provided.

What would you say about a municipality where half the circulating police-cars were secretly fakes "to deter crime", and which would not respond to any emergency?

P.S.: I think as time goes people will be less-accepting of fake-cameras, because they'll assume the operator has a social or legal obligation to provide footage in certain circumstances. For example, to see the license-plates of a hit-and-run in the company parking-lot, or the face of the guy who stole a purse at a restaurant.

The way I see it, dummy cameras and real ones have exactly the same pre-crime effect: they both create a sense of security and add to the observed risk associated with committing a crime. With dummies you get really low cost (practically no maintenance) but lose the benefits if and when a crime actually happens. Still net positive, I think.

Like common locks, cameras only really prevent opportunistic crime. And for prevention, dummy is as good as the real deal as long as it's not widely known which is which.

Only if people don't know about the dummies, things tend toward zero sum with awareness of the situation.

This is an actual thing: "ghost units" - empty police cars placed around to stop people from speeding.

http://www.dailycomet.com/article/20130730/ARTICLES/13073995...

There's a construction site on my commute that has two white Crown Vics parked visibly, with spotlights on the driver side turned out so they show. They're civilian, and I assume they're there to get people to slow down.

The sad thing is I've seen plenty of patrol cars _with_ actual officers in them seemingly not taking breaks per se, but still ignoring blatant traffic law-breaking happening all around them.

Like the corner of Van Ness at Market Street where, during rush hour, a car goes straight through a red light about once every 5 to 10 minutes. I used to cross that intersection everyday and there was typically an officer standing on the southwest corner. I twice told the officer standing there about how frequently cars blow through that intersection and they both admitted it was a problem but had a "nothing we can do about it" attitude. I was flabbergasted. Some pedestrian is going to die there one day.

Several deaths within a couple of blocks:

http://sfgov.maps.arcgis.com/apps/OnePane/basicviewer/index....

Wow. Very interesting. I was surprised how many of those killed are elderly. I wonder if it is because they are more likely to be involved in an "accident"* or as likely to be involved but less likely to survive.

* I use the word accident in quotes because of stories like these: http://www.nytimes.com/2013/11/10/opinion/sunday/is-it-ok-to...

I have seen LAPD park decoys in the lot around big box stores, presumably to deter shoplifting.

I wouldn't say anything to the municipality if it was shown to be a cost-effective measure to improve the situation. It sounds like you presuppose there is also a lack of emergency response capacity, but that is a separate issue.

> What would you say about a municipality where half the circulating police-cars were secretly fakes "to deter crime", and which would not respond to any emergency?

That'd probably depend on a) whether or not they're deterring crime and b) whether or not the overall response time is reasonable.

Well there is a well known psychological principle that a watchful eye prevents crime..

http://www.smithsonianmag.com/smart-news/putting-eyeballs-bi...

I believe that if someone is willing to murder you in front of what they thought was a real camera, you might as well have had a real camera.

many cameras /everywhere/ are 'decoys' although often through lack of maintenance and being broken rather than intentionally.

as much as they are being a security hole, they are also a cheap deterrent. i'd rather we have them than not.

most criminals are smart enough to not want to be seen... despite the misconception that many criminals are stupid, desperate or somehow abnormal people who do not behave like 'the rest of us'.

in the cases where people are genuinely out of their minds it won't make a difference. in that case its much sadder that there is no possible deterrant imo, than that we can not then track down the perpetrator of the crime.

Indeed... cameras need to be maintained when they are broken, the video quality of cameras is not really good for space reason so the footage is not usable all the time, the videos need to be stored somewhere with a good backup system and must obey local laws about data retention, different cameras are managed by different people so you need some paperwork to access everything which might be needed, they need to be checked regularly to see if they still work fine... Putting all of this in place is not really cheap and easy to maintain all the time, that's why not everything works 100% of the time.

Here's the photos. Just because a few weren't working (or real) doesn't mean they have no footage:

http://www.sfgate.com/news/article/Photos-released-of-suspec...

sure, they mention in the article that they have photos, but the message i got from the article is that having the cameras be real would somehow be of benefit because they help you track down the criminals.

to me, it seems very doubtful that the argument presented has any merit. they are described as a 'security gap' in the title and repeatedly in the article, but the evidence supporting that viewpoint is entirely absent.

(i should clarify that its not a worthless argument, just the example used to present it contains evidence only for the counter-argument as far as i can see)

That wasn't taken from a train camera however.

I ride BART every day and have always wondered if the train cameras are even in use most of the time. I suspected not, and this news story confirms my suspicion.

Surveillance is not security, surveillance is surveillance. I an glad to hear that not all cameras monitor our every step.

Dummy cameras feel like 'Security Theater' to me.

But unlike, say, extensive airport screening procedures, it's fairly easy to test if dummy CCTV cameras have an impact on crime rates.

I don't know if such a study has been done, but if it works and impacts the intended endpoints, it isn't really theatre.

I agree with you that there may be a placebo effect.

Perhaps I could call them part of the "set" on which the theater conducts its act.

I wouldn't quite say that. Dummy cameras give petty criminals a risk they hedge their actions against. Security theater would be telling people they can feel safe from personal injury because of said cameras. Someone else put it succinctly: Surveillance is not Security.

You don't have to tell someone that something exists for their safety for it to qualify as theater. I agree that criminals hedge a petty crime vs. getting caught. I now think of this as a sub-type of theater, one that both fakes security and deters incidents via placebo in equal fashion.

You've changed my mind.

No, 'Security Theater' is for when you perform a bunch of actions to make people 'feel' safer without actually making them safer. We accuse airport security screening of being security theater because it is unlikely to be effective against a targeted attack, the very failure mode that the policies were instituted to prevent.

In this case, however, the intended targets are presumably crimes of opportunity, not exhaustively preplanned crimes. For these, the threat of "you're being watched, and if you do something, we'll see it" is probably sufficient. Also they provide low grade confrontation situations to be defused by someone pointing to the camera and saying "There's a camera monitoring this".

Putting up a dummy camera sure seems like one action taken to make people feel safer to me. It isn't the whole theater, just part of it.

Public transportation of the terrestrial variety has very little theater, and very little security anyway. Even the cameras that do work probably exist for the playback, not for the live "surveillance". I just don't think they're watching closely enough.

More panopticon, maybe?

I assume in a panopticon the guards watch the prisoners. I don't think I can assume that about a dummy camera.

In the panopticon, you could be being watched at any moment. If you don't know which cameras are dummies...

There's a big difference between surveillance in public spaces, and private surveillance (effectively unwarranted search & seizure) of citizens. Why would I care if a camera is monitoring me in a public train?

Surveillance can absolutely add security.

How do you propose improving security in this case?

What about paying people to provide security services?

Eyewitnesses might not be foolproof, but they're better than a cost effective dummy camera after a murder has taken place.

I don't think the murderer in this case was deterred at all, perhaps an intimidating rent-a-cop would have allowed the murder to occur near a working security camera where the police could have gathered useful information about the crime.

Does the security need improving?

Same principal we had for cameras in our school buses. Only some of them actually had cameras in them but we were scared of all of them. If the goal is to deter low-level crime like vandalism, I would think they probably worked, at least temporarily.

Although part of me can't imagine thinking someone is watching 4 cameras on every car on every train at all times.

Although part of me can't imagine thinking someone is watching 4 cameras on every car on every train at all times.

Having video footage of crimes is useful even if nobody watches it until after the fact. There are some security cameras in Vancouver which were installed with the explicit promise that footage would be deleted after 30 days without anyone having seen it unless the police get a warrant in the intervening period.

Agreed. I was just thinking if their justification would be to deter graffiti they would want people to be afraid of being picked up during/right after the act, but I guess it would make sense for them to be worried about someone reviewing the footage later and their picture being passed around to the transit cops.

its kind of sad. i'd rather there was a promise that the footage was kept indefinitely, but never watched unless there was a police warrant. that sounds much more useful...

For one, if the data still exists, you can expect its entire history to eventually get datamined via facial recognition because "think of the children" or "we need to identify people's routines so we can spot outliers" or "the terrorists might have taken a train 12 years ago," and eventually used for god knows what else. Not a fan of that idea.

And two, can you actually trust the transit people to keep a pile of petabytes of footage secure forever? I certainly don't. So we should assume that the person datamining all of those records will actually be an adversary.

Third, who's going to pay for it? Video (or even stills) at a useful resolution take up space.

That sounds like the kind of thing that no public agency has the budget for. Even at relatively low resolution, X kb/s * Y cameras * 60 * 60 * 24 adds up to a whole lot of data accumulating every day. Why store all of that data forever when it's vastly unlikely to ever be used after the first 30 days. Even 30 days is probably generous. How often are police not going to have any idea that a crime was committed in the general vicinity of a camera for an entire month, much less two or three or more?

hard disks are cheap. they are not the expensive part of this at all...

In england you know they're really cameras because the feed displays on a screen and it flicks through all of them.

(I'll take some photos tomorrow.)

Yeah. I would imagine that if you had real cameras, it would make sense to publicize that fact. Partially to let people know they shouldn't expect any implication of privacy and partially to let people know the cameras aren't decoys.

The cost/size requirements have come down so much in the past 20 years that it makes a lot more sense to actually install cameras and record locally to a DVR/NVR in the bus.

really? i'm pretty sure i live there and this is the exception rather than the case. particularly on public transport i have never seen this... although i've seen it in restaurants and shops quite a few times, but certainly much less frequently than i have seen cameras, or things that look like them.

They seem pretty common on London buses.

https://www.youtube.com/watch?v=FTHFSuPirXY

never noticed it on them... at least the only one i regularly take.

It doesn't mean that they are actually recording though.

That's why they have that inane message on a loop that says "Security alert, cameras are no gurantee against criminal activity in the BART system."

I've always suspected this. It's very unlikely that video would cross cars, there are already enough pins on the couplers and I believe this would invoke FRA requirements on the engineering.

I'm not sure when the cameras were installed, but the solid-state DVRs usually used on vehicles probably weren't available yet, given the age of the BART rolling stock. Tape-loop recorders would have been very high maintenance.

My suspicion is that some or all BART cars actually are equipped with tape-loop recorders, and what they call 'dummy cameras' are simply recorders which are broken or without tape because of lack of maintenance. The replacement tapes are also getting hard to obtain.

There are several good commercial options for vehicle surveillance now (Safety Vision being a prominent one), but across BART's ~650 cars this would easily turn into a several million dollar project. Let's just keep holding out for the new rolling stock...

Three people were shot in cars on I-80 in Richmond on the same day that this person was shot on a BART train. There are as far as we have been made aware no photos or videos of the event or the perpetrator. Strangely the press is ignoring this other "security gap".

This argument leads to the idea that all decoys are security gaps. Yet it's quite the opposite. No matter how many cameras you have, it's not a gap to have decoys. Because decoys are always cheaper than adding cameras, and security is always resource-constrained.

The grocery store I work at has dummy cameras. I suspect this is because the construction crew broke the security camera system, and nobody is willing to pay for fixing it.

A lady's purse was stolen, and we couldn't do a thing.

I'm still pissed, months after the fact. The manager didn't tell the lady that our cameras are broken, and he only told the officer once the customer was out of earshot (if at all), so the poor lady's probably not aware of the real problem. She could probably sue.

As supplement to right to defend self, maybe, but to replace right to defend self: no.

Doesn't surprise me. Getting stuff certified for trains, cars or planes is expensive. The less of the stuff you have to install, the better, from a financial POV.

High quality dashcam with a month long of looping storage would cost about $50-100 each - locked inside that security department.

Time for them to make them all real cameras.

When purchasing a camera system in NYC for my mom's shop, the tech at the surveillance shop told me that any camera with a red light on it is a fake.

Let's hope that those cameras were supposed to be fake, rather than invoiced and paid for as real cameras but fake ones installed.

I feel for the family that lost a loved one and it's unfortunate they don't have more information about this suspect.

I'm really curious as to how cameras are supposed to prevent a person from killing someone, unless they're doing it in some extremely impractical manner of course.

Reply on news.ycombinator.com