Securing ATMs 24/7
ATMs have become a cornerstone of day-to-day life for millions, but they can also be vulnerable to attack. The global ATM Industry Association reported an increase of ATM crime of 12% for 2017*. And attacks, of course, often involve ATM users, potentially injuring them and causing trauma. But since, by definition, ATMs are often situated outside buildings and used at all times of the day and night, securing them is a challenge – and banks are turning to more intelligent solutions.
An outdoor vulnerability
According to ATM business Association (ATMIA), there area unit associate calculable three.5 million ATMs within the world serving those in would like of money 24/7, one year a year. And there’s wherever the problem in securing them lies. To be effective, most ATMs ought to be publicly areas and open all hours. This, including the very fact that they hold money, makes them a attractive target.
Criminals target these machines – or a lot of exactly – the folks mistreatment them in various ways that. Distracting customers at ATMs in an effort to require their cards or money from them, or to find their PIN number to be used at a later date, for instance. They additionally attempt to install false card readers that steal customer’s card details for use.
On the opposite aspect of the story, banks ought to cope with client disputes. a really tiny share of ATM group actions end in things wherever the client queries whether or not the transaction has been allotted as requested, or perhaps disputes creating the withdrawal in the least. So, security precautions for ATMs area unit a crucial a part of a money institution’s overall security resolution.
Deep Learning technology steps in
Inside associate ATM machine, 2 covert cameras area unit put in, one trained on the user, the opposite on the ATM panel. Deep Learning technology embedded within the ATM security system will find any ‘abnormalities’ within the facial scene before of it, bearing on existing knowledge patterns. So, if there’s another face within the image (for example somebody trying over a user shoulder), or if the person carrying a mask, associate alarm is triggered within the security center.
Using a similar technology, the protection system also can flag if the quantity pad is roofed with a strip to steal PIN codes, or if a false card reader (or ‘skimmer’) is gift to steal card details.
All of those ‘smart’ alarms contour the protection observance method, which means that security personnel will react to period of time situations and not waste time on false alarms. The footage will give proof for any investigation.
The Deep Learning ‘engine’ here would be the Hikvision DeepinMind NVR, that takes the knowledge from the camera and analyses it mistreatment Deep Learning algorithms. this will additionally ‘manage’ footage, in conjunction with alternative NVRs and a video management system, that brings this a part of a complete resolution along with all the opposite components, providing a robust toolset for security and business intelligence.
Securing assets at the extremities of your security resolution could be a heap easier with Deep Learning technology. Even ATMs outside your building is safer, avoiding fraud and protective your customers each day.