I remember reading somewhere that after a certain amount of time it is pointless to have a guard sitting there staring at video display monitors. Essentially they get fatigued and miss stuff. Does anyone know where to find valid information on this subject?
I have a customer where one decision maker wants to employ live monitoring and another wants to avoid it because they think it will be ineffective and possibly expose them to liability if something is missed by a guard.
Thanks.
Brian, this is an excellent question, unfortunately I have not seen a genuine study that confirms this. There are a lot of 'sources' that state this matter of factly but without reference. So you get variants of operators can not watch more than X channels of video for more than Y minutes without missing events, etc. But you do not get a source to a real controlled study that proves this.
To be clear, I think there is general validity to the concern. Also, I don't think the choice is between live or not live monitoring. Event or alarm based monitoring is a hybrid that often makes sense.
I am not sure the "guard staring at monitors" scenario really still exists. Live Monitoring is typically event driven these days, where events can be alerts from an access control system, motion on cameras, analytics, etc. Guards are not "sitting there" just staring at monitors, events are put in front of them (via VMS or automation software).
A key to making live monitoring work is that it needs to be a full-time job for the operator. To add complexity, you are not likely to have events 100% of the time, so there needs to be other tasks for the operator that can be easily interrupted/paused when an event comes in.
For smaller operations, it is difficult to strike this balance, which is why the remote live monitoring operations are becoming more common. They can staff up enough, and fill operators time with other tasks, to make the model more workable.
Is your customer debating hiring their own person that would be on-site in some way, or hiring a remote monitoring company?
"I am not sure the "guard staring at monitors" scenario really still exists."
As much as I hate to say it, I can assure you it does still exist. As much as I push analytics, alerts, and integration with other systems to focus on responding to events nothing seems to sway customers from having static video walls with a guard or two watching them. It is the single biggest frustration I have with clientele. In a way, I feel it is that clients want a picture worthy NASA operations center to show off.
Even more frustrating - a $80k video wall with one dispatcher who is also checking in/out keys, responding to every call into the center, entering items into their guard management software, etc.
I agree with this. I've seen entire state prison systems, schools, airports, cities, and probably any other vertical you can think of in the past year which still just have people sitting and watching video. Full time.
In some of those cases, maybe there are "events" as in someone calls with an issue, but I'd say the vast majority are still not event driven. It's sad and probably futile, but that's still really incredibly common.
I can also confirm it. We have a customer that employs live guards to watch monitors and walk the grounds. Though they have reduced the guards from 3 shifts down to just 1 at night, it still happens. I wonder if event driven video has reduced the need though.
This is actually for a public high school. The guards would be in house security personnel. They would be dealing with visitors in addition to watching the monitors.
I agree with the event driven model, we successfully use this for perimeter protection. However I am not a huge believer in the analytics that would be required to catch problems in a high school hallway. So I am not sure that we could depend on events, which leads me to the conclusion that live monitoring would be mostly ineffective?
In the scenario you described I would agree live monitoring would be of questionable value. Unless the school is the kind of environment where things are occurring regularly, like 4+ times/day, the guards will get bored watching the video, or will do other things instead.
I also agree that analytics are a tough sell in schools for daytime applications when classes are in session. You could program rules to alert on activity in the halls only when class is supposed to be active. That will still likely generate a lot of events from teachers who have free periods, maintenance staff, etc., but would help to make the guard aware of activity, even if they just use that alert to pay peripheral attention to the screen.
I have seen this statistic, but can't find the actual research to back it up. It is often quoted in articles, presentations, etc...
Research in the US has shown human observers start showing signs of viewing fatigue after as little as 12 minutes, overlooking up to 45% of all activity in the scenes. After 22 minutes, they overlook up to 95%.
Robert Moore, citing a 2002 study published in Security Oz magazine in his article 'What are the latest trends in intelligent video and analytics?' SP&T News, June 6, 2008. See:www.sptnews.ca/index.php/Editorials/Ask-the-Expert-Smarter-AI/menu-id-106.html
The link is of course long dead...
This thread reminded me of this...
Who else saw the Moonwalking Bear on the thumbnail and was looking for it in the video?
Brian,
There was a study or research done. I recall seeing the study at one point a few years ago. I don't recall but I think it was a study out of the UK. If I locate it I will share a link to it or include a pdf.
Avigilon published a white paper on the topic in 2014 called "Attention Economics
Whitepaper".
The Paper referenced the following documents.
Robert, Thanks!
Here are the links to those reports:
How effective is human video surveillance performance?
Visual sustained attention: image degradation produces rapid sensitivity decrement over time
Lastly, The Appropriate and Effective Use of Security Technologies in U.S. Schools, from 1999, it is more of a general (outdated) guide on using cameras, etc., e.g., wayback claims of 'Currently, the cost of a color camera as compared to an equivalent blackand-white camera is anywhere from 30 percent to 70 percent greater.'
Granted humans can miss something but what are the limitations of analytics?
Can analytics determine if there is group of people and a fight breaks out, an assault, a car break in, someone pointing a gun?
I am really not that up to speed on analytics, so is this possible?
I am really not that up to speed on analytics, so is this possible?
I have installed 120 cameras at a local hospital that highly depends on their camera viewing to enhance their lack of manpower. I have to say, they have been very successful and have seen a decrease in thefts (etc.) due to monitoring. They also have the duties of answering the phones etc. as well. I think the fact that they are encouraged to use the cameras to patrol the parking areas keeps the staff alert and attentive.
Here is a NIJ study about video viewing.
#7, Thanks. Here is the key relevant quote from the NIJ study:
Experiments were run at Sandia National Laboratories 20 years ago for the U.S. Department of Energy to test the effectiveness of an individual whose task was to sit in front of a video monitor(s) for several hours a day and watch for particular events. These studies demonstrated that such a task, even when assigned to a person who is dedicated and well-intentioned, will not support an effective security system. After only 20 minutes of watching and evaluating monitor screens, the attention of most individuals has degenerated to well below acceptable levels.
It is important to note that study was published in 1999, which means the study they cite was from 1979.
There is a genuine question whether the camera quality of 1979 (low res, probably b&w only, etc.) impacted the level of attention / effectiveness vs 2017 cameras.
Some slightly more recent studies from the UK tend to bear out the "vigilance decrement" in CCTV monitoring. One example is here. While this is an academic postgrad thesis it cites a whole lot of other studies from 1998 to 2008 ... eg.
In a field study performed as part of the URBANEYE project by (Norris and McCahill, 2006), CCTV operators were observed while performing their tasks. The results revealed that only 35% of the total detection was due to proactive monitoring. Furthermore, according to Velastin et al., (2006), after 20-40 minutes of active monitoring, CCTV operators will suffer from “video-blindness”, and cannot recognise objects on the video anymore.
[Velastin in turn refers to "research carried out by the Police Scientific Research Branch at the UK’s Home Office (Wallace and Diffley, 1998)"]
The UK Government's CPNI guidelines for CCTV monitoring control rooms still refer to a "the vigilance decrement, which typically occurs after 20 to 30 minutes of continuous work (varying with the level of concentration required)" - however they reference this from the same 1983 study John linked to above...
Interesting research more recently suggested (in a limited study) that there is no vigilance decrement for surveillance "specialists" - although more common in novices and generalists, and also there's still a lot of issues with what is missed...
Conclusion out of all of this is that the degraded monitoring performance is clearly an issue for most people, but that not many researchers have investigated it since the 90's and are still citing back to a small number of research papers.
From the Abstract portion of the study you linked to:
"Vigilance decrements were found for novices and generalists, but specialists maintained their performance for the first hour and then increased it."
You referenced 'no vigilance decrements for surveillance specialists' - but the study actually goes further, stating that these surveillance specialists performed better after they monitored cameras for more than an hour.
I find that conclusion to be highly suspect - and most probably either a flaw in their testing methodology, or the study was commissioned to find that conclusion (by someone who profits from advancing this theory; i.e. monitoring stations that employ 'surveillance specialists' to stare at camera feeds instead of using analytics to do it faster and more efficiently).
Ask questions and get answers to your physical security questions from IPVM team members and fellow subscribers.
