The very nature of the systems we develop for customers in all the markets we serve — aerospace, space, ground transportation, security and defence — has given Thales a head start as data volumes explode. In all these markets, we have found specific ways to make sense of our customers' data, intelligent ways to create value by turning that data into information they can act upon.
In a world of mobile devices and social media, the volumes of data being generated today are quite phenomenal — an estimated 90% of all the data on earth have been generated in the last two years! This relatively recent development has shaped the marketing strategies of Internet giants such as Apple, Google and Amazon, and created a huge amount of media buzz around the idea of Big data. But the Internet of Things is taking the phenomenon to a new level. Road sensors, GPS receivers, smart electric meters, and a whole host of interconnected objects and wearable devices, are already making their entrance, generating data volumes of a different order of magnitude and creating a much coveted potential for turning that data into actionable insights.
Thales first got involved in the Big data revolution in 2009, when we were part of a joint R&D project to detect and investigate cases of online payment fraud in France. With 10-15 million transactions a day, the volumes of data to be processed were enormous. Added to that, decisions needed to be made extremely quickly — in less than 300 ms — to determine whether a transaction was fraudulent or not. And, because of the highly relational nature of the data to be processed, with 40 million cardholders and 500,000 online merchants, we quickly realised two things: that traditional database models were inadequate; and that new processing algorithms would be needed to leverage the potential of distributed computing.
In every market we serve, Thales designs its systems to keep pace with the relentless growth in demand for higher performance and greater security. These systems are built around large numbers of components, including video cameras, satellites, chemical sensors, electronic tags and so on, each of which generates unprecedented amounts of data. Handling the burgeoning load of data calls for new storage and processing technologies (NoSQL, Hadoop, Cassandra, etc.) as well as innovative algorithm solutions (such as MapReduce, parallelisation or linearisation).
The Thales approach is structured around three closely interconnected themes: Big data (data management/storage), Big Analytics (processing, enrichment and value creation) and Visual Analytics (exploitation and interactive visualisation of data).
Based on this three-pronged approach, Thales develops specific business cases with each customer, defining exactly what they expect from the process of collecting, storing and analysing data. The Thales approach opens up completely new decision-making opportunities, particularly when decisions are time-sensitive and based on complex information.