WEPMR055 Proceedings of IPAC2016, Busan, Korea openSE: A SYSTEMS ENGINEERING FRAMEWORK PARTICULARLY SUITED TO PARTICLE ACCELERATOR STUDIES AND DEVELOPMENT PROJECTS* P. Bonnal †, B. Féral, K. Kershaw, B. Nicquevert, CERN, 1211 Geneva 23, Switzerland M. Baudin, École Normale Supérieure, UPMC & CNRS UMR 7203 LBM, 75005 Paris, France L. Lari1, Fermilab, Batavia IL 60510, USA J. Le Cardinal, École Centrale/SupÉlec, 92290 Châtenay-Malabry, France 1also at European Spallation Source, 221 00 Lund, Sweden Abstract practices, the project management corpus differs substan- Particle accelerator projects share many characteristics tially depending on the professional domain it is applied with industrial projects. However, experience has shown to. One can typically observe different approaches for the that best practice of industrial project management is not eight following domains: construction, process industry, always well suited to particle accelerator projects. Major new product development, new service development, differences include the number and complexity of tech- information and communication technologies, organiza- nologies involved, the importance of collaborative work, tion, events and human resource development. Research development phases that can last more than a decade, and projects shall be considered separately as they do not the importance of telerobotics and remote handling to completely fulfil all of the generally agreed definitions for address future preventive and corrective maintenance projects in organizations. So-called scientific projects requirements due to induced radioactivity, to cite just a differ from research projects in the sense that they are few. The openSE framework it is a systems engineering aimed at providing means for performing research pro- and project management framework specifically designed jects. In many research domains, scientific projects share for scientific facilities’ systems and equipment studies many of the characteristics of industrial projects. This is and development projects. Best practices in project man- the case for particle accelerator facilities that have many agement, in systems and requirements engineering, in engineering aspects in common with process industry telerobotics and remote handling and in radiation safety facilities, such as chemical plants or power stations. management were used as sources of inspiration, together While “reasonable-size” projects belong typically to one with analysis of current practices surveyed at CERN, GSI of the eight domains mentioned above, large-scale scien- and ESS. tific facilities are composed of sub-projects from all eight domains [2]. INTRODUCTION The conception, development and construction of large- scale scientific facilities rely on the appropriate use of The conception and development of large-scale scien- project management practices. However, these practices tific facilities emitting ionizing radiations rely more on are not unique; they are many and specific to certain as- project management practices in use in the process indus- pects of the project. Forcing all project contributors to try than on systems engineering practices. This paper implement a unique project management approach must aims at briefly highlighting possible reasons for this pre- have a rationale. Sharing a common core is a prerequisite sent situation and to propose a way to enhance systems for enhancing communication and coordination among engineering so that the specific radiation safety require- project participants, although the definition of this is not ments are considered and integrated in the approach. We straightforward. Some suggest the implementation of have reviewed lessons learned from the management of PMI’s Project Management Body of Knowledge [3], large-scale scientific projects, and more specifically that which aims to describe project management practices of the Large Hadron Collider project at CERN. It is suited to all types of projects. The recently released ISO shown that project management and systems engineering 21500 Std. [4] “can be used by any type of organization, practices are complementary and can beneficially be including public, private or community organizations, and assembled in an integrated and lean managerial frame- for any type of project, irrespective of complexity, size work that grants the appropriate amount of focus to safety and duration”. Both documents are general-purpose and radiation safety aspects. standards, but are not sufficiently specific to fulfil the SCIENTIFIC PROJECT MANAGEMENT expectations mentioned above. For that very reason, they can serve as a basis for the core framework, but cannot be Scientific Project Specificities the core framework. While the NASA’s Systems Engi- Even if some professional associations such as the US neering Handbook [5], or the ESA’s ECSS Standards [6], Project Management Institute [1] aim to homogenize propose such a core framework for space projects, to our knowledge, nothing similar exists for large-scale particle ____________________________________________ * Work supported by European Commission, FP7 (G.A. no. 264336). accelerator projects, and more broadly for scientific pro- † email address: [email protected] jects. ISBN 978-3-95450-147-2 07 Accelerator Technology Copyright ©2398 2016 CC-BY-3.0 and by the respective authors T33 Subsystems, Technology and Components, Other Proceedings of IPAC2016, Busan, Korea WEPMR055 Systems Engineering dedicated to this subject [12]. Safety is also scarcely in- troduced in IEEE Std. 1233 [13] (p.9-10). Textbooks According to the International Council on Systems En- related to systems engineering are many [11], [14], [15]. gineering (INCOSE) [7], “systems engineering (SE) is an Of the few reviewed, all of them mention safety as an interdisciplinary approach and means to enable the reali- important requirement for the development of a complex zation of successful systems. It focuses on defining cus- system. For instance, Sage and Rouse consider safety as tomer needs and required functionalities early in the de- part of the “scientific and engineering effort” of SE, as velopment cycle, documenting requirements, and then transverse activities beside “reliability, maintainability, proceeding with design synthesis and system validation survivability, human engineering, and other factors” [11] while considering the complete problem” [8]. In other (p.13). It is also worth mentioning that none of the re- words, SE can be seen as a subset of the project manage- sources cited above introduces the two equivalent con- ment corpus dedicated to the development of complex cepts of ALARA (As Low As Reasonably Achievable) or mechatronics systems. ALARP (As Low As Reasonably Practicable). Several academic studies confirmed that the implemen- tation of standardized practices increases project success In conclusion, it should be highlighted that systems en- (see for instance Milosevic and Patanakul [9]). But other gineering complements project management practices by studies convey the contrary, such as that of Dvir, Raz and providing means to communicate and coordinate the Shenhar [10]. To these authors: “The findings suggest technical dimension of the project. However, while sys- that project success is insensitive to the level of imple- tems engineering is particularly well suited for the con- mentation of management processes and procedures, ception and development of complex products that are which are readily supported by modern computerized made of subprojects of a mechatronics nature (mechanics, tools and project management training. On the other hand, electronics, controls software), it has not been designed project success is positively correlated with the invest- for complex facilities. ment in requirements’ definition and development of technical specifications”. The SE corpus is somehow in THE openSE FRAMEWORK line with this conclusion: it suggests that particular atten- tion be paid to the technical side of the project. What Is openSE? The ORAMS Requirements openSE is the systems engineering framework that CERN and a few particle physics research institutes have Out of the four concerns of ORAMS (Operability, Reli- worked out to streamline project management processes. ability, Availability, Maintainability and Safety), safety More specifically, it is intended to provide means to effi- may impact the success of a project quite differently from ciently manage development projects of complex systems the others because of the way its deliverables are as- subject to or emitting ionizing radiation, while paying sessed. While the four ORAM requirements are associat- particular attention to four important aspects that may ed with tangible deliverables, the “safety success” of a otherwise be omitted because of the natural focus that is system relies on intangible deliverables. If, over its deve- naturally given by the project team to the project deliver- lopment and operations, there is no impact on people, no able itself. accident, and no impact on the environment, then a pro- ject is considered a success. However, in terms of ORAM openSE Governing Principles requirements, whose assessment is based on something Five principles governed the development of openSE, happening, the “safety success” will not be seen because namely: openness, leanness, participation, modularity and “nothing” tangible has happened. For this reason, stake- scalability [16] (pp. viii and ix). holders are usually keener to work out the ORAM side of Openness.