Enabling distributed computing in swarms of cyber-physical systems
Host organizations
Hiring Institution
École Nationale de l’Aviation Civile (ENAC)
PhD-Awarding Institutions
École Nationale de l’Aviation Civile (ENAC)
The University of Adelaide (UoA)
Position Description
Download and read the full position description before applying
Proposed Projects
Option 1
Exploring the Design of Distributed, Fault-Tolerant Computing for Nanosat Constellations
The recent deployment of an increasing number of nanosatellites in low-earth orbit (LEO) presents new opportunities for space applications. Built atop small-sized yet powerful blocks, known as CubeSats or simply nanosats, nanosats constellations emerge as promising platforms for massive sensing and large-scale distributed computing. Indeed, they represent a cheaper, competitive alternative for traditional satellite systems for a wide range of application domains such as earth observation and defence.
However, the design of distributed, intelligent systems based on nanosats is particularly challenging: nanosats have more stringent physical limitations with respect to processing/networking capability, energy supply, and connectivity among nanosats.
Therefore, this project aims at investigating distributed systems problems and propose specific solutions for dynamic reconfiguration mechanisms, consensus algorithms, and data replication schemes on nanosats systems.
This project will be supervised by researchers from both ENAC and UoA, and it will be conducted in close cooperation with SpaceLocker company that will contribute with its invaluable experience in nanosat systems.
Option 2
Designing Fault-Tolerant Middleware Services in Swarms of Cyber-Physical Systems
Cyber-Physical Systems (CPS) enable complex applications to solve real-world problems, and as a research field, they are rapidly evolving. The architecture of a CPS strongly relies on the tight coupling between the cyber components (i.e., hardware, software, including control models) and the interactions with the physical environment (basically through actuators and sensors). In contrast to traditional computer systems, the efficiency and reliability of cyber components depend not only on the platform and the inter-process communication, but on the intrinsically unpredictable interactions with the environment, as well.
This doctoral project aims to provide distributed, fault-tolerant computing techniques to foster the design of reliable applications for emerging CPS platforms. In order to face the important challenges to build such techniques properly, this project adopts a data-centric approach to design middleware services with both correctness and performance guarantees, focusing on key features including fault tolerance, Interoperability and flexibility.
This project will be supervised by researchers from both ENAC and UoA, and it will be conducted in close cooperation with SpaceLocker company that will contribute with its invaluable experience in nanosat systems.
Option 3
Towards Reliable Distributed Coordination for Mobile Edge Computing
Recent years have seen rapid technological evolution in embedded systems, that have become easier yet cheaper to design and deploy. As a result, new, larger embedded platforms and experimental facilities are emerging, in particular atop swarms of Unmanned Aerial Vehicle (UAV) and constellations of small satellites. These advances brings new opportunities for the design of novel services, such as accurate natural disaster forecast from huge amounts of Earth observation data or multi-tasking robotic testbed for lunar scientific missions.
To successfully reach the aforementioned ambitious design goals, these platforms should be capable of operating reliably, at scale and, eventually, meeting real-time constraints. Yet the lack of suitable distributed protocols for coordinated, concurrent task execution is likely to hamper the efficiency of these powerful platforms, leading to undesirable waste of resources, costly operations and unpredictable performance guarantees.
Therefore, we argue that novel reliable protocols, based on shared memory, reliable communication and replication state machine (RSM), will play a key role in emerging embedded systems. For instance, we believe that RSM is a promising technique to provide challenging functionalities of new control system for robotic platforms, including autonomous, reliable multi-task execution with transparent fault tolerance.
Our goal in this doctoral project is twofold. First, we aim to study the fault tolerant techniques for reliable distributed coordination of multiple autonomous robots. Second, we aim to investigate the opportunities to improve the reliability of multi-task planning and control of emerging robot platforms, in particular swarms of drones and constellations of nano-satellites.
Supervisors
Research Areas
Computer science, Robotics, Electronics, Embedded systems