It’s well known that automobiles and smartphones have several sensors to improve today’s capabilities over those of previous generations. However, they are not the only products that impact consumers’ lives to do so. For example, automated teller machines (ATMs) have a surprising number and different types of sensors to perform a variety of functions. Even though the number of ATMs (475,000 – 500,000 units) deployed in the U.S. is small compared to cars and phones, consumers expect them to work when they need them. Sensors play a critical role in making them work properly.
In a technical report from 2015, sensors were identified as essential items in smart ATMs for diagnostics, security and environmental concerns. As shown in Table 1, the diagnostic sensors enable the basic functions of the ATM. In contrast to sensors that allow normal operation, the sensors in Table 2 identify situations that can cause security alerts to be activated in abnormal situations. Table 3 sensors identify proper operating range for the environment.
Table 1 – Sensors for Diagnostics
- Cassette Wheel Sensors
- Stack Empty Sensors
- Thumper Position Sensors
- Shuttle Entry Sensors
- Cam Sensors
- Gate Entry Sensor
- Gate Cam Sensors
- Gate Media Sensors
- Hall Sensors
- Shuttle Media Sensors
- Transport Motion Sensors
Table 2 – Sensors for Security
- Smoke Sensor
- Intrusion Detection Sensors
- Door Position Sensor
- Floor Motion Sensor
- Shock Sensor
- RF Signal Detection Sensor
Table 3 – Sensors for the ATM Environment
- Pressure Sensor
- Temperature Sensor
- Humidity Sensor
- Battery Health Check Up Sensor
Sensors that activate a security alert like the Floor Sensor that triggers an alert immediately if the ATM is being tilted or dragged away from its actual position or the Shock Sensor that provides a notification when drills or hammers are being used for unauthorized entry appear to be rather obvious and implemented early in ATM designs. In contrast, RF Sensors that detect RF signals to find fraud through skimming are a more recent addition.
Detecting fake currency is a universal ATM issue. In the U.S., the size and weight of paper currency are tightly controlled. The seven US bills are all 2.61 inches wide by 6.14 inches in length and 0.0042 inches thick. Each bill weighs 1 gram. A variety of schemes are used to identify fake currency including: a portrait watermark, security strip, micro-printing (very tiny lettering in various places in the note), color-shifting ink, raised printing and a 3D security ribbon or thread that is actually a hologram. Any one or all of these visually identify real bills and they can be used in sensors for machine verification.
In India researchers, used a stack of tactile sensors that consists of 19 or more light sources and optical / laser / UV / IR sensors to provide a histogram from a reference point to identify fake bills.
Facial recognition and fingerprints allows user identification without a password. They just need to be compared with and match the faces and/or fingerprints in the database. In June 2021, CaixaBank in Spain introduced ATMs with facial recognition. The software and cameras were capable of identifying over 16,000 points on the user’s face.
References
Microsoft Word – ATM Industry Demographics and Issues.docx (atmia.com)
(PDF) Smart ATM Machines (researchgate.net)
Detecting Counterfeit Money – Carnation Bill Money Counting Machines (carnation-inc.com)
IAM : Integrated Active Monitoring Pvt. Ltd. (smartiam.in)
(PDF) Money to ATM -Fake Currency Detection (researchgate.net)
