A number of human controlled systems operate 24/7 (24 hours per day, 7 days a week). Examples include transportation and surveillance systems, or military operations. 24/7 system level properties, such as the weekly schedules provided by an airline, arise from the human-operated coordination and control of a set of resources which typically do not have 24/7 availability, due to operational constraints such as maintenance. These properties persist over time and space, independently of the resources used to guarantee their validity.
With recent advances in unmanned vehicles, we are able to form a networked vehicle system (NVS) that can be deployed to perform 24/7 persistent operations in autonomous fashion, i.e. without the intervention of human operators (with the exception of maintenance activities). However, the state-of-the-art in NVS is still far from providing tools and technologies to design an autonomous NVS with 24/7 persistence in a systematic manner and within an appropriate scientific framework. This is because we lack models and control methods for systems of dynamic networks of interacting vehicles. In these systems, information and commands are exchanged among vehicles, and the roles, relative positions and dependencies of those vehicles change during operations.
The main objective of this project is to develop modeling and control methodologies for a NVS in persistent autonomous operations.
We build on the experience of the research team in the design and deployment of NVSs for oceanographic studies, air traffic management, transportation, telecommunications, and sensor networks.
The research team, including researchers from the Faculty of Engineering from Porto University (FEUP), the University of California at Berkeley (UCB), the Politecnico di Milano (PoliMi) and the Royal Institute of Technology (KTH), has fielded unmanned air, ground, surface and underwater vehicles in innovative operations in Europe and in the United States.
We plan to develop an inter-disciplinary approach that builds on advances in (1) dynamic networks of hybrid automata (DNHA); (2) hierarchical architecture design for semi-automated, distributed teams of agents; (3) incorporating human intervention in mission planning and execution.
The approach will be tested and evaluated on a case study concerning the 24/7 surveillance and environmental data collection for the Leixões Port Authority (APDL). We will use several Light Autonomous Underwater Vehicles (LAUV) equipped with multi-beam sonars and environmental sensors, the Swordfish Autonomous Surface Vehicle (ASV) and drifters equipped with environmental sensors. Testing and evaluation of our developments both in simulation and in a demonstration will be facilitated by our software toolset, which includes the Neptus command and control framework, the Seaware real-time publish-subscribe middleware and the DFO/Dune control framework.

The project is organized into six tasks:

Task denomination Start date End date Duration
T0. Project management 14-06-2010 13-06-2013 36
T1. Case study 14-06-2010 13-12-2011 18
T2. Simulation framework 14-08-2010 13-08-2012 24
T3. Models 14-12-2010 13-06-2012 18
T4. Control methodologies 14-03-2011 13-09-2012 18
T5. Physical Demonstration 14-06-2012 13-06-2013 12