THAAU —  Integration of Industrial Devices   (13-Oct-11   08:30—10:15)
Chair: S. Perez, CEA, Arpajon, France
Paper Title Page
THAAUST01 Tailoring the Hardware to Your Control System 1171
  • E. Björklund, S.A. Baily
    LANL, Los Alamos, New Mexico, USA
  Funding: Work supported by the US Department of Energy under contract DE-AC52-06NA25396
In the very early days of computerized accelerator control systems the entire control system, from the operator interface to the front-end data acquisition hardware, was custom designed and built for that one machine. This was expensive, but the resulting product was a control system seamlessly integrated (mostly) with the machine it was to control. Later, the advent of standardized bus systems such as CAMAC, VME, and CANBUS, made it practical and attractive to purchase commercially available data acquisition and control hardware. This greatly simplified the design but required that the control system be tailored to accommodate the features and eccentricities of the available hardware. Today we have standardized control systems (Tango, EPICS, DOOCS) using commercial hardware on standardized busses. With the advent of FPGA technology and programmable automation controllers (PACs & PLCs) it now becomes possible to tailor commercial hardware to the needs of a standardized control system and the target machine. In this paper, we will discuss our experiences with tailoring a commercial industrial I/O system to meet the needs of the EPICS control system and the LANSCE accelerator. We took the National Instruments Compact RIO platform, embedded an EPICS IOC in its processor, and used its FPGA backplane to create a "standardized" industrial I/O system (analog in/out, binary in/out, counters, and stepper motors) that meets the specific needs of the LANSCE accelerator.
slides icon Slides THAAUST01 [0.812 MB]  
THAAUST02 Suitability Assessment of OPC UA as the Backbone of Ground-based Observatory Control Systems 1174
  • W. Pessemier, G. Deconinck, G. Raskin, H. Van Winckel
    KU Leuven, Leuven, Belgium
  • P. Saey
    Katholieke Hogeschool Sint-Lieven, Gent, Belgium
  A common requirement of modern observatory control systems is to allow interaction between various heterogeneous subsystems in a transparent way. However, the integration of COTS industrial products - such as PLCs and SCADA software - has long been hampered by the lack of an adequate, standardized interfacing method. With the advent of the Unified Architecture version of OPC (Object Linking and Embedding for Process Control), the limitations of the original industry-accepted interface are now lifted, and in addition much more functionality has been defined. In this paper the most important features of OPC UA are matched against the requirements of ground-based observatory control systems in general and in particular of the 1.2m Mercator Telescope. We investigate the opportunities of the "information modelling" idea behind OPC UA, which could allow an extensive standardization in the field of astronomical instrumentation, similar to the standardization efforts emerging in several industry domains. Because OPC UA is designed for both vertical and horizontal integration of heterogeneous subsystems and subnetworks, we explore its capabilities to serve as the backbone of a dependable and scalable observatory control system, treating "industrial components" like PLCs no differently than custom software components. In order to quantitatively assess the performance and scalability of OPC UA, stress tests are described and their results are presented. Finally, we consider practical issues such as the availability of COTS OPC UA stacks, software development kits, servers and clients.  
slides icon Slides THAAUST02 [2.879 MB]  
How Many Marriages Last 10 Years? A tale of Industrial Relations for Accelerator and Experiment Controls at CERN  
  • R.J. Jones
    CERN, Geneva, Switzerland
  Funding: The work of the CERN openlab project is funded jointly by CERN and the openlab industrial partners: HP - Intel Corporation - Oracle – Siemens.
The CERN environment is highly reliant on industrial control systems such as control actuators, remote profibus Input/Output modules, PLCs and SCADA systems. The LHC accelerator and experiments are examples of large-scale industrial plants and, as such, are of great interest to commercial companies. CERN and the LHC collaborations work closely with ETM, the provider of the SCADA system, and Siemens for PLCs and other devices within the context of the CERN openlab project. For nearly 10 years the CERN openlab project has been bringing together leading companies to evaluate and integrate cutting-edge technologies and services. The CERN openlab project is a highly visible example of a successful public-private partnership. Through close collaboration with these companies, CERN acquires early access to technology that is not yet commercialised. In return, CERN offers expertise and a highly demanding environment for pushing new technologies to their limits and provides a neutral ground for carrying out advanced R&D. This presentation will explain why the CERN openlab project came into existence and how it has evolved over the last decade. Examples of the technical achievements from the control domain will be highlighted. The motivations and expectations of CERN as a research organisation and the industrial partners as commercial companies will be explored as well as the experiences gained and lessons learnt. Through this presentation, the audience will gather insights into how research organisations could build their own win-win model for industrial relations.
CERN website:
CERN openlab website:
slides icon Slides THAAUIO03 [1.984 MB]  
Keynote. Customized COTS Technologies Through Industry – Research Facility Partnership  
  • J.T. Truchard
    National Instruments, Austin, USA
  The current economic climate has put even more focus on keeping projects under budget and on time while using the latest technology to meet the needs of measurement, diagnostic, and control systems. Commercial off-the‐shelf (COTS) systems take advantage of innovations in the computer industry and the hundreds of millions of dollars devoted to R&D – domain experts can now benefit from FPGAs as wells as multicore CPUs and GPUs without being specialists on these technologies. Through collaborations between industry and research facilities, engineers can customize these technologies while keeping costs low to achieve faster computing and loop rates. With every project lasting 15 to 20 years, obsolescence management is yet another key benefit of industry‐research collaborations. At this session, examine the technological and business benefits of this type of partnership.  
slides icon Slides THAAUKP04 [7.502 MB]