Feed: An advanced method of detecting bank robberies has been developed by Portuguese scientists, building on experience they have gained over the creation of space flight simulation software.
Feedzai's software is used to monitor each online marketplace in Portugal, while global media controls payments worth $ 229 billion a year.
Space missions and rogue detection systems have one thing in common: the need for highly sophisticated software. "When a spacecraft launches, it needs software to guide it," explains Paulo Marques, who was ESA's adviser before setting up Xenum X Feed. "You also need software to communicate with the ground."
Long before a boat launches, the software needs to be thoroughly scrutinized for problems. However, there is a problem, as Paulo points out: "you still do not have the spacecraft."
So, scientists are building a "universe" of software to simulate the mission. "What you need is something that depicts the spacecraft, mission control and ground stations, as well as many other elements to be able to control everything."
In the ESA, Poulau and Nuno Sebastiano of Feedzai used high-performance computing techniques to create virtual satellites: "computer bugs that pretend to be everything involved. A computer works as a spacecraft. "
The software must be very stable to be able to imitate perfectly all elements of the mission and board, and must be able to do so quickly. This software is also used for operator training.
Corresponding challenges are also addressed in fraud detection: for example, both large-volume data processing is required in real-time, and the software must be able to recognize anything moving outside of its normal sphere.
However, a significant difference is that in the case of space missions, there are some fixed and non-existent rules, while fraudulent detection has to be decided on a case by case basis: a temperature change in a spacecraft is always a problem, however each bank customer implies different habits and data .
Therefore, software should be able to detect what is normal for an entrepreneur and what about a teacher before showing "strange" behavior.
Feedzai has developed an artificial intelligence system that can process a large volume of transactions by checking every single transaction made by a customer within four years. Using machine learning and large data techniques, the software "learns" to distinguish between "suspicious" and "non-suspicious" transactions.
Feedzai software monitors over 300 variables per person, generating very detailed and personalized expense profiles. "In total, we are monitoring over five billion variables at once," says Paulo.
"It's like having 500 highly intelligent people track each individual transaction and decide on the basis of their experience if it is a fraud or not. It's a huge amount of computing power, "he adds.
According to Carlos Serquero, of the Instituto Pedro Nunes, technology to save money for banks while improving customer loyalty: "Feedzai's machine learning models and big data science are able to detect 30% frauds earlier in relation to traditional methods and demonstrate how the capabilities developed in ESA research centers can be useful in other areas as well ".
According to the ESA's announcement, 11,4 billion dollars each year are lost due to credit card fraud - a number that is expected to rise as cybercriminals are developing their methods.