New technologies for watching the coast ...

Canada has the longest coastline of any country in the world (243,042 kilometres/151,019 miles) and insufficient means to guard it all. Maintaining a presence in the Arctic, and providing security along the coast, is a challenge (some would call it a nightmare): tasks include providing surveillance, relief for vessels in distress, security management for events (such as the upcoming 2010 Winter Olympics in Vancouver), responding to natural disasters, and search and rescue.

Coastal and arctic-wide area surveillance and response is managed from central command and control centres. They provide a composite situational view by collecting, processing, storing, and displaying data generated from operations. Decision-making in these centres occurs in a cycle of Observe-Orient-Decide-Act (the OODA loop, or Boyd cycle). An entity (either an individual or an organization) that can process this cycle quickly has a higher chance of meeting its objectives. For the agencies involved in coastal safety and security, this requires the use of advanced decision support tools to be done well.

This summer, the CanCoastwatch (CCW) consortium will wrap up its work. As noted here, multiple organizations have collaborated on the development of a system to support command and control centers for wide-area surveillance and response.

The littoral surveillance problem is characterized by the use of mobile and fixed surveillance resources over a large geographic area in order to identify, assess, and track the maximum number of moving, stopped or drifting objects. Surveillance and tracking capabilities are scarce, and this makes it very difficult to perform large volume surveillance and to keep track of all activities. Hence, applying the mobile resources in the best possible manner increases the coverage quality and the likelihood of a successful operation.

Two key requirements for command and control systems for wide area surveillance and response are:

• Situation assessment: Whether it’s managing routine surveillance, or responding to events, information needs to be rapidly extracted in order to provide a good operating picture. This extraction and information fusion is done from multiple sources and transfered over less-than-perfect communication channels. The stream of information fusion provides mission controllers with frequent situational updates that enable new insights.
• Decision proposal: The decisions that have to be made are all about how to best utilize available resources. Situation assessment is likely identifying changes in the operational environment that call for dynamic reallocation of the available resources, as well as resources that are already deployed in response to some event. Efficient deployment of multiple search resources is a challenging task for humans, since it involves handling a large number of constraints and optimizing within dynamically changing parameters.

The CCW consortium’s main objectives were:

• To create an advanced simulation test-bed for the purpose of evaluating the effectiveness of network-enabled operations in coastal wide area surveillance situations;
• To develop new data fusion algorithms to provide improved situational awareness, and
• To develop new methods for rapidly reallocating resources to meet strategic and tactical objectives.

The test-bed that we’ve put together provides a flexible multi-agent architecture that allows researchers to test distributed data fusion and dynamic resource management algorithms in a setting that is realistic enough to draw valid conclusions about the effectiveness of network-enabled large area coastal surveillance applications. The test bed includes a multi-layer plug-and-play architecture, and represents a general framework based on John Boyd’s OODA loop, supporting a variety of realistic constraints on resources and their communications networks.

Actenum’s responsibility in the CCW consortium was to develop the resource allocation components. The best strategy for resource management and scheduling in such dynamically changing environments has, up to now, been poorly understood. Traditional methods in AI and Mathematical Programming assume a static and non-distributed problem, which does not apply to littoral surveillance and response. We used our constraint reasoning technology platform to develop new distributed and dynamic resource management techniques. These are used for resource deconfliction, load balancing, and resource migration, and to plan and schedule resource deployment and rapidly redeploy in reaction to new information.

More reading

1. Canada Coast Watch (CCW) Final Report, Precarn Incorporated, March 31 2001
2. A Testbed simulator to evaluate the efficiency of net-enabled surveillance, H. Wehn, R. Yates, A. Dlugan, D. Clark, C. Nadeau, H. Zwick, E. Bosse, A. Guitouni, P. Valin, M. Irgens, B. Havens, P. Harvey, R. Farahbod, U. Glaesser, H. Leung, Beyond the Next Frontier, UVS Canada 2006, Montebello, Quebec, 7-10 Nov. 2006.
3. Application of Search Theory for Large Volume Surveillance Planning, A. Guitouni, K. Jabeur, M. Allouche, H. Wehn, J. Happe; To be published in Proceedings of Fusion 2008: 11th International Conference on Information Fusion, Cologne, Germany. July 2008.
4. Network-Centric Sensor Resource Management, B. W. Johnson, J. M. Green, 7th International Command and Control Research and Technology Symposium, 2002.
5. Toward Practical Knowledge-Based Tools for Battle Planning and Scheduling, A. Kott, L. Ground, R. Budd, L. Rebbapragada and J. Langston, Proc. IAAI 2002.