Control Sheet No. 19

A Turnkey Control System for ELI-NP

By : Gašper Pajor (Cosylab)

Cosylab as a part of the EuroGammaS consortium wins the tender to build and deliver world’s most powerful gamma beam facility.

The Extreme Light Infrastructure (ELI) [1] currently consists of four projects that will provide a great platform for the study of the fundamental processes that unfold during light-matter interactions. The three facilities currently on the way are ELI-Beamlines near Prague, Czech Republic, ELI-ALPS (Attosecond Light Pulse Source) in Szeged, Hungary and ELI-NP (Nuclear Physics) in Magurele, Romania. The site of the fourth project (ELI-Ultra High Field) is still to be decided.

ELI-NP [2, 3] will consist of two large machines, namely:

  • A very intense (1013 γ/s) gamma source with a brilliant γ beam, a 0.1% bandwidth and with Ev > 19 MeV. This is obtained by incoherent Compton back scattering of laser light off a very brilliant, intense, classical electron beam (Ee > 700 MeV) produced by a warm linac.
  • A very high intensity laser, where beams from two 10 PW lasers are coherently added to get intensities of the order of 1023 - 1024 W/cm2 and electrical fields of 1015 V/m.

Once built, the ELI-NP will be the most advanced laser and gamma beam facility in the world.

The gamma source machine will be delivered by the EuroGammaS consortium which is headed by the Italian Institute of Nuclear Physics (INFN). Institutes and companies from all across Europe are participating in this consortium and Cosylab assumed responsibility for delivery of the overall control system.

In a complex system like the ELI-NP, a unique control system will manage the whole machine, from the gun to the radiation beam-lines. This means that the control system must be able to execute commands on all the active elements and control all the diagnostic devices, giving the needed information to the accelerator operators. Furthermore, it must be easy to upgrade the system - substituting old elements or introducing new ones.

The EPICS framework was chosen for the ELI-NP control system, as requested by the customer. The implementation will be based on a modular control system design that leverages solutions developed at other similar facilities. This along with a set of strict processes and conventions ensures that any custom, ELI-NP-specific development will be done in a standardized and coherent manner.

Along with the whole set of technical challenges, the ELI-NP also faces two mutually amplifying risks; the timeline is very tight, as the first stage of installation and commissioning is due in 18 months and at the same time the stakeholders are geographically dispersed across 8 European countries. The control system is heavily exposed to both of these risks; we will interface with practically all other systems and most, if not all, of the control system will already be needed to be up and running in the commissioning phase. It will definitely be a challenge for the Cosylab ELI-NP team!


Figure 1: Artist’s impression of the ELI-NP facility that is being built in Magurele, Romania. [3, 4]


Figure 2: Representatives of the EuroGammaS Consortium on 19 March 2014 in Rome, Italy, after the signing of the contract to implement the Gamma Beam System within the ELI-NP project.




Gašper Pajor is a Group Leader at Cosylab, joining in 2001. He has a background in physics, with extensive experience in project management of control system projects. Currently, he is the project leader for the ELI-NP turnkey control system, that is responsible for delivery of the overall control system for this project. In his free time, Gašper enjoys spending time with his family in his house in the countryside they just moved into.

Reusing Expertise: A Control System for Solaris

By : Frank Amand (Cosylab)

The first Polish synchrotron radiation facility Solaris is being built at the Jagiellonian University Campus in Kraków and Cosylab is selected to realize the control system.

More than a decade ago, Polish synchrotron radiation users formulated the ambition of a Polish national light source facility. Together with 35 research institutes and universities, they founded the Polish Synchrotron Consortium. The project, named Solaris [1], received financing from the European Structural Funds, and is being constructed as we speak, with first research planned for 2015.

Specific to this project is a strong partnership with the MAX-IV project [2] in Lund, Sweden. The design of the synchrotron ring aims at maximizing reuse of the design of the 1.5 GeV storage ring of MAX-IV. This has repercussions on the design of the control system: it is based on the TANGO control system [3], chosen by MAX-IV. Reuse of Tango Device Servers, developed for use at MAX-IV is envisioned and is a major assumption of the control system integration project. If successful, it will strengthen the cause of international, multi-site, open-source collaborations, such as TANGO controls.

The novelties in storage ring technologies, such as a single iron block design for the DBA’s and NEG coated vacuum system are illustrated here [1] as well as the typical beam parameters.

Cosylab has been selected for the delivery of the control system integration services, including delivery of the timing system hardware. With this choice the Solaris team has opted not to reinvent the wheel on control system integration, just as they chose not to reinvent the storage ring design. Instead they leverage timing system expertise built-up with projects like the MedAustron medical accelerator [4] in Austria. Similarly control system integration expertise for rings and beamlines, e.g. the booster synchrotron of the Australian Synchrotron [5] and a Turnkey Control System for the PHASE Beamline at the Kurchatov Centre for Synchrotron Radiation [6] in Russia.


Figure 1: Artist’s impression of the Solaris facility being built at the Jagiellonian University Campus in Kraków, Poland. [1]




Frank Amand, Belgian, joined Cosylab in 2011. Previous work experience includes 12 years with Royal Philips in Belgium and the Netherlands in a variety of software engineering related roles. His technical expertise lays in the domain of human-computer interaction, GUI design and usability. He is currently Cosylab’s Head of Marketing.

FECOS: Front End Controller Operating System

By : Samo Tuma (Cosylab), Miha Vitorovič (Cosylab), Matej Klun (Cosylab) and Mathias Eichinger (MedAustron)

What is FECOS?

FECOS is a light-weight, purpose-built, programming framework that allows users of the MedAustron control system to write control code for the accelerator front end controllers (FECs). One of the key purposes of FECOS is the standardization of FEC interfaces and partially behavior, e.g. FECOS enforces a state machine that is implemented by every FEC. FECOS also provides standard functionality like error reporting, logging, timing system support and watchdog and host health-status monitoring, providing a mechanism for configuration, control and supervision. FECOS runs on all FECs and each FEC is able to run multiple applications (application components) [2-4].

FECOS and all device-support software are implemented through object-oriented programming in LabVIEW, also known as LVOOP. This approach established a streamlined development process with MedAustron developers [3].

How does FECOS work?

All communication is event based and FECOS itself is state machine based. Figure 1 shows the FECOS class hierarchy.

All device control is implemented by extending a Component class. Each Component within the framework is provided with a pair of queues that are used for asynchronous inbound and outbound communication. Executive is the main dedicated component that tracks and manages the “user applications”, i.e. components that are statically bound to the framework. It also takes care of communication – grabs components outgoing messages and dispatches them to the appropriate target components.


Figure 1: FECOS Class Hierarchy [4]

How is communication handled?

FECOS was originally conceived to standardize operation and communication of the front end controllers. The framework implements a few communication protocols (for example, HTTP and MAPS (MedAustron Publisher-Subscriber)) [3].

The application components that run on the FECs are not concerned with the details of the different communication protocols. These are handled by communication protocol wrappers, which are exposed to the developer and provide a layer of abstraction to the actual interface. Each component within the framework is provided with a pair of queues that are used for asynchronous inbound and outbound communication via messages. Messages are called events and each event has a type or meaning and carries additional information that the receiver of the event can use to perform its tasks. The framework handles inbound and outbound communication asynchronously to the application components.

Each application component implements a general purpose state machine for uniform control procedures. The framework always contains a dedicated application component, the “executive” that tracks and manages the “user applications” that are statically bound to the framework. [3]

A user relies on a set of FECOS libraries which contain LabVIEW VIs (virtual instruments) and data structures to develop applications. [3]

While upstream communication to the processing tier and other peers at the same tier is transparently handled by the framework, direct access to hardware and software elements is not performed through a dedicated API. The programmer uses the functions and components provided by National Instruments or third-party providers for this purpose. [3]

In conclusion

Presently FECOS is successfully running on a variety of devices along the MedAustron accelerator structure. These devices include ~300 magnets and their corresponding power converters, RF systems, ~150 beam diagnostics devices and ion sources. Most of these devices are also synchronized by the Main Timing System FECOS component.

Currently Cosylab engineers are participating in work on the Medical Front End which will result in additional LabVIEW components built on the FECOS framework. The aim of the Medical Front End is to ensure complete safety for patients being treated at MedAustron.

FECOS is a good example of the continuous collaboration between Cosylab, MedAustron and National Instruments’ engineers which resulted in a modern state-of-the-art Front End Controller Framework, the first one ever implemented purely in LabVIEW. The Front End Controller Framework is now being finalized with feedback from MedAustron engineers.


Samo Tuma joined Cosylab in October 2012 as a software engineer, then spent most of his time working on the MedAustron project. His current role is that of a senior software/hardware engineer, interested in a good game of basketball or a bite into a chilly pepper!

Miha Vitorovič is a Senior Software Developer at Cosylab. He has a background in Computer Science and is currently Project Manager for the Data Management system for ESS, that is responsible for delivery of the Machine configuration, naming, cabling and lattice applications. In his free time, Miha enjoys spending time with his family, hiking and enjoying the mountains.

Matej Klun joined Cosylab in February 2011 as a software/hardware engineer. Currently he is in the role of a senior systems architect, but when there is time left he enjoys a pint of his own home-made beer.

Mathias Eichinger joined MedAustron in November 2012 as a Control System Expert. He is currently managing the extension and maintenance of the front end controller software developed in LabVIEW.



  1. J. Gutleber, A. Brett, R. Moser, M. Marchhart, C. Torcato de Matos, J. Dedič. The MedAustron Accelerator Control System, Proceedings of ICALEPCS2011, held from 10-14th October 2011 in Grenoble, France. (Paper: (Presentation: (
  2. R. Štefanič, R. Tavčar, J. Dedič, J. Gutleber, R. Moser. Timing System Solution for Medaustron; Real-Time Event and Data Distribution Network. Proceedings of ICALEPCS2011, held from 10-14th October 2011 in Grenoble, France. (
  3. Matej Šekoranja, FECOS Overview.
  4. Matej Šekoranja, Miha Vitorovič, Rok Štefanič, FECOS,

GUI Style Guides: Haute Couture Window Dressing?

By : Frank Amand (Cosylab)

Experimental psychology research on visual perception has taught us that color perception is a very complex matter. Managing the color use in GUI’s is NOT a matter of Window Dressing, it is a vital element to good Usability.

A professionally made GUI style guide will not just give your developers a head-start and make the GUI’s look better: it is a matter of actual improved functionality and reduced operator error!

Compare the two screens on this page. I grabbed the top one from a Google image search for “control room GUI” [1]. The image below is a “makeover” of the first, following the rules of a style guide. The makeover was done by simply replacing the colors of shapes and backgrounds in Photoshop and making small design adjustments: no Botox, no surgery, just make-up and eyeliner! ;-)





Note: My purpose is to make a few general points in this article and not to make a judgment about this particular screen, its author or purpose. This screenshot happened to be readily available and very illustrative to the article.

So, let’s ask ourselves a few questions:

If this GUI wants to tell us about an important problem with the machine, in the form of an alarm, in which of the versions would you spot it easiest?

A traffic light on Time Squares in backlight vs. one in the suburb at dusk?

That’s why the style guide for the made-over GUI would tell you:

  • Use only neutral greys (with their RGB values) for panel backgrounds
  • The background is then subdued and screens look consistent.
  • Use such and such preset colors that are Lightness values (tones) of the same blue Hue (and no other hues of blue) for the main panels on the screen: titles, groupings, lists, …
    • This is aesthetically pleasing and uncluttered: lets the active, foreground elements stand out
  • Reserve the use of highly saturated colors for important predefined functions such as alarm levels. Avoid large patches of these colors and have a boundary around them to separate then from the background
  • For similar functions across screens, use the same control with the same appearance. For example, do not change the appearance of a toggle button without a very good reason.

At the top, the style guide would say a few things that would make the general layout of the screen look different in the first place:

  • Leave enough white space between elements and between elements and the border. At least 10 pixels.
  • Group elements that belong together using “Gestalt principles”
  • Put them close together
  • (Strictly) Align them

If your team of 2-3 people has to produce say 50 screens, talking to many stakeholders for requirements, design iterations and you give them “carte-blanche” in terms of this pure look-and-feel aspect, what are the chances they come up with a consistent set of screens?

Indeed, a style guide with “keep-it-simple” rules, a few applicable GUI design patterns and some examples will energize your team and overcome the blank page syndrome rather than stifling their creativity.

Does a style guide solve all GUI design problems? No! A cosmetic makeover of a screen with real usability flaws will not fix those. For that you need to do a good interaction (re)design exercise. But the user experience (UX) will be affected if 50 screens have the same consistent, restful, unconfusing look. It’s a great first step to improving control screens.

And to get back to the provocative “Haute Couture?” in the title.

Should good, consistent GUI styling be seen as merely a “Haute Couture” exercise?

Well, doesn’t the original screen more resemble John Galliano’s creatures then the “makeover”? [2]

Compare them with those “cuties in blues and grey” next to them. [3] Which ones look more helpful running your accelerator? ;-)




  2. Christian Dior Autumn-Winter2008, John Galliano


Frank Amand, Belgian, joined Cosylab in 2011. Previous work experience includes 12 years with Royal Philips in Belgium and the Netherlands in a variety of software engineering related roles. His technical expertise lays in the domain of human-computer interaction, GUI design and usability. He is currently Cosylab’s Head of Marketing.


The author would like to thank Andreas Luedeke from PSI for his valuable feedback on a draft of this article.

The Photo Board

CS19-Hiroshima Uni CSL Tshirt wi

Cosylab T-Shirts On April 14th, our President Mark Plesko and Vice-president for Asia, Prof. Emeritus Shin-Ichi Kurokawa, visited Hiroshima University. In classical Cosylab style, Mark presented a lecture about accelerators and gave a T-shirt for each correct answer from the audience. A particularly inspired postgraduate student put on his T-shirt immediately and thus got the honour to be “photographed with our President”. Those who know Mark and his casual dress style may wonder what is more surprising: that a usually timid Japanese student put on a Cosylab T-shirt or that Mark is wearing a suit and a tie. But in Japan, everything is possible.

CS19-BNCT Takashi Mark

Cosylab Japan Develops a Turnkey Control System for the BNCT project During a visit by Mark Plesko to the Cosylab Japan office in Tokai, developer Takashi Nakamoto put on the signature Cosylab jacket for a “partner-look” style photo in front of the BNCT accelerator, under construction in Tokai. If you have read the article on GUI Style Guides, then you can expertly judge for yourself whether the jacket colours match the coil windings and contrast with the yokes - or should I say jokes? Oh yes, and no suit here this time.”


Sponsorship in Action Cycling in Ljubljana is an extremely popular mode of transport. The House of Experiments (Hiša Eksperimentov) took this idea and created a fun interactive experiment at the recent “Znanstival” - cycling across the Ljubljanica River! Since Cosylab was the main sponser of the Znanstival, staff had a special opportunity to cycle across the river. Pictured is Rok Tavčar proudly flying (or maybe hanging!) the Cosylab banner!

back to previous content

Download printable (pdf) version of Control Sheet no.19 here.