Science and Technology meet operational needs
MSA supports effective and efficient decision making and enables maritime operations to preemptively identify emerging safety, security or environmental issues so that a timely intervention is possible. MSA highly depends on the ability of sensing, collecting and processing technologies to handle the big data challenges brought by the ever increasing volume, velocity and variety of data, which often lack veracity.
Information relevant to the operation is collected from networked sources (sensors and non-sensors), and processed to provide an ideally robust, complete, and coherent maritime picture. Additionally, MSA tasks aim at establishing patterns of life, against which anomalies can be identified. The maritime picture correlated with other background information supports the detection, tracking, assessment, and prediction of illicit activities as well as safety-related events.
Understanding the maritime situation enables decision makers and emergency responders to focus on relevant events, to prevent malevolent acts, to minimize the impact of a possible threat, and/or to intervene in a timely manner. To reach a common and comprehensive understanding of the maritime operational environment, accurate, timely and standardized information need to be shared among nations, partners and civilian agencies, providing the required information superiority to successfully conduct maritime operations.
From coastal radar to satellite imagery, developments in Intelligence Surveillance and Reconnaissance (ISR) technology promise complementary and possibly persistent surveillance capabilities to build the maritime picture. This includes sensor technology deployed onboard Maritime Unmanned Systems (MUSs), such as Unmanned Surface Vehicles (USV) and Unmanned Underwater Vehicles (UUV). These have recently demonstrated high performance in terms of scalability, adaptability, robustness, persistence and reliability. Deployed within a monitoring region, MUSs can cooperatively form heterogeneous intelligent networks to detect, localize, and classify targets. Opening up to new operating scenarios, MUS introduce also new scientific and technological challenges on autonomy, distributed and collective intelligence and sensing, data fusion, detection and tracking.
At the front-end of sensor systems (radar, sonar, camera, etc.) is signal processing, which builds mathematical models to investigate the principles of detection, localization, tracking and classification of targets. By associating and combining measurements of different sensors, data fusion techniques help to build coherent and clear picture of a region of interest wherein multiple objects appear, move and disappear.
At the same time, in the era of big data, processing infrastructures provide an unprecedented capability to gather, store and process massive amounts of data in real-time. In the maritime domain, sources are heterogeneous and provide data in many formats (structured or unstructured); moreover, data are often intermittent, sparse and noisy. Computational scalability and data models allow cueing together all the available data, discovering (otherwise hard to find) patterns.
While humans can leverage from their experience to distinguish between normal and anomalous patterns, such interpretation and contextualization capabilities are everything but immediate for machines. Artificial Intelligence (AI) aims at mimicking some human cognitive abilities to automate routine tasks in an efficient way. Machine learning focuses on the learning process from large amounts of data, to further perform classification or prediction tasks. Supervised machine learning, such as deep learning techniques, together with unsupervised approaches, such as clustering, have recently demonstrated that significant improvements are possible in the maritime domain (e.g., ship classification, maritime traffic characterization and pattern recognition).
Other AI approaches can support understanding by bringing transparency, interpretation, or explanation to reasoning and decision making, addressing cognitive tasks such as vessels behaviour analysis or intent assessment. Additional to automated reasoning and uncertainty handling, knowledge representation can provide harmonized and rationalized terminology of the maritime domain, required for information sharing environments or integrating solutions toward technical, procedural and human maritime interoperability.
To improve human-machine teaming, knowledge acquisition techniques can bring the human “in the loop” in the early stages of the system design. Capturing knowledge and know-how in a structured and reproducible way helps in the understanding of specific aspects, such as procedures of maritime operations, information needs, and cognitive bias.
In this perspective, the achievement of MSA requires a multi- and interdisciplinary approach, spanning several fields including, but not limited to sensing technologies, signal processing, unmanned systems, data fusion, machine learning, big data, artificial intelligence, and applied human factors.
Objective
Focusing on Maritime Situational Awareness, the objective of MSAW 2019 is to bring together scientists, engineers, researchers from scientific communities with national and international authorities, end users and operators, and industrial representatives. MSA specialists will present and discuss scientific and operational challenges, advanced technologies and knowledge gaps, in order to facilitate future collaboration and research activities. MSAW 2019 encourages contributions from EU H2020 projects, as well as other research initiatives, to present on-going progress, results, and/or live demos.
FOLLOW US: