The CanCoastWatch project

In November of 2005, Precarn provided nearly $1 million in funding to the CanCoastWatch (CCW) project, which brings together a team of researchers from industry, Canada’s military, and academia, to create an advanced simulation test bed to evaluate the effectiveness of network-enabled operations in a coastal surveillance situation. Of particular interest to the project participants was the challenge of maintaining the efficient deployment of multiple search resources, and rapid extraction and distribution of key information.

CCW is designed with a flexible, multi-agent architecture that allows researchers to test dynamic resource management, distributed data fusion, and dynamic resource management algorithms in a realistic setting. This allows valid conclusions to be drawn about the effectiveness of network-enabled large area coastal surveillance applications.

The challenges of large-area coastal surveillance

The Canadian government faces a number of challenges in maintaining effective surveillance of the coast, whether Arctic-wide, or East/West. Mobile (maritime patrol aircraft, helicopters, unmanned aerial vehicles, ships) and fixed (for example, land-based radar) surveillance assets are deployed over a large geographic area to identify, assess, and track as many moving, stopped, or drifting objects as necessary in a specific situation. The observed objects are not necessarily aware of being observed and may be cooperative or non-cooperative, as well as friendly or hostile. The scarcity of surveillance assets (such as electro-optical, infrared, and synthetic aperture radar sensors) and tracking capabilities (normal radar modes) make it very difficult to perform large volume surveillance and to keep track of all activities.

Managing communications, data fusion, resource management, and scheduling in such a dynamically changing environment requires advanced decision support capabilities. Traditional methods in AI and Mathematical Programming assume a static and non-distributed situation, but in this case, neither assumption is valid. The resource scheduling problem dynamically changes as the environment and requirements change, due to continual information updates. Moreover, not all information is always available to all parts of the network: there are communication and computing delays, bandwidth constraints, and communication losses to consider.

With multiple surveillance platforms available at any given time, there is a need to network them, and keep each current as the situation evolves. The initial resource/platform deployment creates a dynamically improving awareness picture of the situation, while ancillary information is continually supplied from external sources; this contributes to the evolution of the situation assessment.

The project concluded successfully in the summer of 2008.