How can complex, globally distributed data sets be efficiently stored and used for integrated data analysis? An end-to-end solution from T-Systems provides the answer: signal data is compressed by up to 90 percent without loss of information and stored in edge devices. The data is processed in an optimized big-data format, resulting in the processing time of the data being 40 times faster than the standard procedure. Engineers send analysis requests to the edge clusters distributed around the world for calculation. Only the results – i.e. the relevant data – are transmitted via secure connections, which leads to significant time savings, since the big data can be used for ad-hoc analyses. Measurement data is enriched with metadata during processing, so that engineers can search for the data relevant to them semantically and easily.
The big data & global edge analytics lowers costs for companies and increases the scope for processing and working with large amounts of data. With Big Data Signal Processing, raw data are read out on site, compressed without loss of information, and prepared for analysis. They remain in the edge device, a kind of mini data center on site. The edge devices are part of an integrated database which is available to companies through all cloud storage systems. Globally distributed development teams submit semantic analytics queries that are distributed worldwide via federation for calculation. Only the analysis results flow back, which significantly reduces response times and uses less bandwidth.
How can companies efficiently process complex, globally distributed data sets and use them for integrated analyses? The automotive development sector is also facing this challenge. When testing vehicles in the snow of Finland or in the desert of Dubai, several terabytes of signal data are collected daily for each prototype. All this information should be available to globally distributed engineering teams for diverse and spontaneous evaluations as soon as possible following the test.
T-Systems’ new end-to-end edge computing solution is setting standards in this area. The innovative software Big Data Signal Processing captures and stores the data without loss of information and up to 90 percent compressed. The data remains where it was created, in a mini data center on site, the edge device.
Here’s the trick: data are no longer transferred to a central cloud for analysis. Instead, analysis requests from engineers are sent via federation to globally distributed edge clusters and calculated locally. The results flow back via a secure network connection. This has two advantages: firstly, parallel analysis queries are possible from any development site and, secondly, this method accelerates data processing by a factor of about 40. T-Systems has already planned and implemented the customer-specific big data infrastructure for a company and operates it across the entire ICT stack from a single source.
Big data and global edge analytics offer added value to all companies that wish to work with large volumes of data in a decentralized manner and for whom speed and cost efficiency are important. For example in the areas of construction, testing, production, or logistics.
