Using ecological concepts to improve software design in Collaborative Adaptive Systems:
The importance of diversity-stability concepts
Postdoctoral position
Scientific Framework
Collaborative adaptive systems (CAS) integrate heterogeneous devices (ranging from sensors to phones), which collaborate through a network of software modules to provide global services such as environmental monitoring or urban transportation control. These large-scale systems operate continuously in open, unpredictable environments and must react to internal perturbations (stress or disturbance) and adapt to environmental dynamics. All these characteristics indicate multiple similarities between CAS and other forms of complex systems.
The scientific objective of this postdoc will be to investigate (i) how models of biodiversity dynamics in ecosystems can be transferred as the theoretical foundation for the establishment of innovative software mechanisms for the spontaneous increase and management of diversity in CAS and (ii) how far the relation between CAS and other forms of complex systems holds.
More precisely, some ecological processes based on the concept of diversity/stability (independently of the scale) may be similar to CAS. We can cite for example 1.) Network organisms (e.g.: clonal plants, ants, …) which modules can specialize locally in order to adapt to heterogeneous environments and promote performance at the network scale. This scale can either comprise one individual (the case of plants for example) in more local conditions or the population (for strongly mobile organisms) at the landscape scale. The investigation of analogies between such structures and CASs, from an ecological perspective will entail research on topics such as decision making processes, information integration, trade-off costs and benefits of plastic variation in response to environmental heterogeneity. 2.) species coexistence at the community scale linked with their functional traits. In this case, species interactions depending on their traits may be the fundamental processes that will result in a better functionality at the community level (food-web dynamics can be included in this example). For these two examples part of the individual, population or communities change plastically in order to adapt to environmental conditions (and especially heterogeneity and predictability) and the appearance of such functional local specialization may generate emerging properties promoting performance at the higher level. In these examples, we focus on the mechanisms that trigger this specialization, how it occurs and what effect it has on the performance at the higher level.
The postdoc fellowship will be funded by the European FET DIVERSIFY project for a period of 3 years starting early 2013. The postdoc will integrate the Landscape and Climate change Research Group at the UMR ECOBIO from the University of Rennes1, France (http://ecobio.univ-rennes1.fr/) and will work closely with the INRIA group leading the project and the other partners of the consortium.
The postdoc’s mission will include the following activities
- survey the literature on different biological processes that present analogies with CAS (we cite two different examples above but many others can be looked at) in order to generate general patterns on the role of diversity in performance relatively to environmental heterogeneity.
- learn about the challenges for software construction in CAS and assist computer scientists for the transfer of biodiversity models
- participate in the definition of precise experimental protocols to evaluate the impact of software diversity on the dynamics of CAS. The design of these protocols will be inspired from protocols used in experimental ecology.
- develop and maintain connections with the ecological board composed of European ecologists advising the DIVERSIFY project about biodiversity models and their dynamics
- participate in the integration and dissemination of the project
Skills
Applicants should have a PhD in theoretical (plant) ecology and strong motivation for pluridisciplinarity. Applicants should have also some knowledge on experimental design in ecology. Because strong relationships will have to be developed additionally within the group of ecological experts involved in the project (Pr. M. Hutchings, Pr. B. Kunin, Dr. E. Thébault, Dr. C. Melian; Dr. J. Baudry and Dr. F. Burel), mobility for short stays in the other research teams in France or abroad is therefore necessary. Knowledge on plant ecology or/and landscape ecology may be appreciated.
Contacts
Interested candidates should send their CV and a statement of research interests to:
Cendrine Mony
Benoit Baudry
Cendrine.mony@univ-rennes1.fr
+33 (0)2.23.23.64.79
http://ecobio.univ-rennes1.fr/fiches_perso/CMony/
Benoit.baudry@inria.fr
+33 (0)2.99.84.72.98
http://people.rennes.inria.fr/Benoit.Baudry/
Review of applications will begin right away, continuing until decision has been made