What is the role of a scientific leader?
The formal description of the role of scientific leader is to evaluate and approve the scientific value of the deliverables and the research directions. My preferred approach is to work as a collaborative facilitator - providing co-ordination of the research activities and steering our team in the right direction i.e. toward the project goals. It is a very satisfying role for me... working in a team environment, sharing knowledge and in doing so I am helping researchers innovate and ensuring their work evolves from scientific publishing to more tangible outputs.
Can you tell us a bit about your background / earlier research and how it complements the CHOReOS project goals?
I began my academic career following a degree in computer science - unable to let go of the desire to learn and acquire knowledge I wanted to find a conduit for my passion. This led me to a role in research which allowed me to create new knowledge in the fields of ambient computing, pervasive computing and middleware technologies.
Fundamentally this work has been the backbone of my research focus on the development and evolution of distributed software. It has been 10 years now that I have been actively researching the area of mobile software, devices and services and my research has included research on efficient solutions to semantic SOA services and pervasive service discovery in a multi-radio, multi-protocol networking environments.
Thus, not only I am able to bring my earlier research initiatives to bear upon the CHOReOS project but also today's research projects such as CONNECT, a European Commission FP7 project, that seeks to define tomorrow's middleware solutions that will alleviate the challenges of interoperability by synthesising on the fly the connectors via which networked systems will communicate. This project is particularly interesting because many of the CHOReOS partners also participate in this project thanks to the collective domain expertise they bring.
Is there existing work in this domain you would like to acknowledge or prior research that will influence the direction of CHOReOS.
It is no coincidence that much of the work already existing in the domain that I would reference is the product of research amongst the CHOReOS partners, being subject matter experts in this domain they have already contributed exemplary work that we will rely upon as part of the CHOReOS project. Thus, it would be fair to say we will be building on our partners' existing research and furthering that in the context of choreographies.
We state that we are re-visiting this area of computing as, for example, a draft specification for choreographies emerged as a proposed web service standard though it did not evolve into a mature specification. With the evolution of computing towards ultra-large scale networks of networks there is a growing realisation that the concept of choreography has returned to the forefront of computing once again.
What are the outputs from the CHOReOS project that you see as being essential-for/integral-to future middleware technologies.
All of them and it's not just middleware! The CHOReOS project outputs will also target development support and requirements engineering - these will be accessed via the CHOReOS Integrated Runtime and Development Environment (IRDE). The CHOReOS project will be disruptive in nature as it will make possible the design of tomorrow's solutions encompassing 'billions of services' i.e. the ultra-large scale future Internet of services.
What are the CHOReOS project's key scientific challenges?
A key challenge for us is that we will target the global ecosystem and this scale is unknown - the evolution of this 'scale' is advancing rapidly and we realise this is a significant challenge for us. Other challenges we will need to face are those of Quality of Service (QoS), thus a large part of our research work looks at Verification and Validation mechanisms (V&V).
With so many services, devices and users how will choreographies manage exceptions and facilitate graceful degradation simultaneously at such scale and complexity?
The development of ultra-large scale choreographies is a complex task and this should address not only standard choreography behavior but also behavior in the presence of faults. Using choreographies does not solve the problem per se although the distributed nature of choreographies may be seen as an advantage regarding fault tolerance.
is "Directrice de recherche" at INRIA. Valérie joined INRIA Paris-Rocquencourt in 1999. Since 2002, Valérie is the head of the ARLES INRIA research project-team at INRIA-Rocquencourt. Her research interests relate to distributed systems, software engineering, pervasive computing/ambient intelligence systems and middleware. Valérie has (co)authored over 100 technical papers in the area of distributed systems and software engineering, and has been involved in a number of European and industrial projects. Valérie has been and is serving as PC member, including as PC chair, of leading international events in the area of distributed systems, middleware, software engineering and trust management. Valérie was the coordinator of the FP6 IST STREP PLASTIC project, currently the coordinator of the FP7 FET CONNECT project and scientific leader of the FP7 ICT CHOReOS project.
CHOReOS is a project of the FP7 European program: FP7-ICT-2009-5 - Objective 1.2 (grant agremment n° 257178)