CERN and Kontron: When Matter matters

LHCb is one of four experiments in the Large Hadron Collider (LHC) – and a very fascinating one at that: The experiment is set up to explore what happened after the Big Bang. “Fourteen billion years ago, the Universe began with a bang. Crammed within an infinitely small space, energy coalesced to form equal quantities of matter and antimatter. But as the Universe cooled and expanded, its composition changed. Just one second after the Big Bang, antimatter had all but disappeared, leaving matter to form everything that we see around us — from the stars and galaxies, to the Earth and all life that it supports”, write the CERN scientists. They are trying to determine the small differences between matter and antimatter to find out why our universe is mainly made of matter and not antimatter. The experiment will continue until at least 2018.

pixel/@fotolia.com

Big Data from Sensors

Its basis is a 4.500 ton detector system with almost one million sensors. These sensors are registering the different particles that emerge while protons, that are moving almost as fast as the light, are colliding in the 27 kilometer long accelerator.  The detector system consists of several sub detectors. Each sub detector is specialized in measuring parameters like energy value or track sequence. About 2.000 events per second are being recorded – resulting in a huge amount of data: 250 GB per hour.
CERN wanted to virtualize the servers that are responsible for controlling this particle accelerator experiment. For this task the research institution was on the look-out for smart nodes which not only had to be very compact, robust and cost efficient but would come with a long term availability and high scalability, too. Today lean 1HE rack systems, each with four processor modules, ensure the efficient and fail-safe processing of the data.

Expensive Experiments need Condition Monitoring

A condition monitoring system ensures the availability of the very complex detection system. This control is necessary to avoid mistakes during the expensive experiments. Only if the condition monitoring registers any anomalies in the components the personal has to interfere manually, otherwise the parameterization of electronics are done automatically.
Instead of accessing the decentralized 1HE server with a complete WinCC SCADA software from the control center, today, thanks to the virtualization, all monitoring functions are managed in one central system. The virtual machines are not only more flexibly adaptable in regard to the different applications: Energy efficiency and availability are much higher as well.

Smart Nodes for Performance and Reliability

Smart nodes have replaced the decentralized servers. Their job is to ensure the safe transfer of condition monitoring data between the detection system and the central virtualized condition monitoring server. The nodes are equipped with x86er intelligence and host an integrated ELMB-OPC server for three ELMB systems each (which are connected via CAN) and an OPC server for the VME systems (connected via CAN as well). 
Each of the four 19 inch 1HE sub systems in one 1HE system chassis contains four independent carrier boards with four computer-on-modules. The reliability of the installation is higher and on-site maintenance was significantly reduced.
CERN decided for COM Express® mini, because this very small form factor ideally suited the compact system design and COM Express is the leading standard for computer-on-modules. The API is delivered by Kontron as well: market leading middleware KEAPI (Kontron Embedded Application Programming Interface) makes access and control of hardware resources in embedded applications much easier.

If you want to learn more about the CERN project you can read the whole story on the German Website computer-automation.de (unfortunately, at the moment the documentation is available only in German).