In the aftermath of the December tsunami, we posted a variety of articles about the idea of a distributed emergency alert system. Such a system should be able to work over a variety of media, without being tied to a proprietary network or format. It should be open, so that it could be modified to meet local needs and new requirements. And -- most importantly -- it should be embraced by existing emergency networks and first responders, and not simply serve as an idealized model. Such a system now exists: the Common Alerting Protocol, or CAP.
CAP is an open, standardized alert information format usable to collect and disseminate warnings and reports of hazards and disasters, natural or otherwise; it's XML-based, so it's usable on a wide assortment of devices and media. CAP notes, documentation and commentary can be found at the wiki-based "CAP Cookbook." The protocol has been in development since 2001, but the tsunami seems to have accelerated its adoption. The April, 2005, draft of the v.1.1 protocol is available online (PDF).
The underlying value of CAP is the standardization of message format:
CAP defines a single message format with the essential features to handle existing and emerging alert systems and sensor technologies. This standard format can replace a range of single-purpose interfaces among warning sources and disseminations channels. CAP addresses the concerns about compatibility and operational complexity that have been stifling development. [...]
A key benefit of CAP for sending alert messages is that the sender can activate multiple warning systems with a single input. Using a single input reduces the cost and complexity of notifying many warning systems. A single input message also provides consistency in the information delivered over multiple systems. People receive exact corroboration of the warning through multiple channels. This is very important, as research has found that people do not typically act on the first warning signal but begin looking for confirmation. Only when convinced that the warning is not a false alarm, do they act on it.
CAP is designed to be compatible with all kinds of information systems and public alerting systems, including broadcast radio and television as well as public and private data networks. Rather than being defined for one particular communications technology, CAP is essentially a "content standard": a digital message format that can be applied to all types of alerts and notifications.
The Office of Management and Budget has enjoined all Federal agencies to implement CAP. The U.S. Geological Survey is leading in CAP implementation for the Department of the Interior. Earthquake notifications in CAP format will be available from USGS by the end of June, 2005. CAP messages for landslides and volcanoes will be available by the end of September, 2005. These CAP messages will also be disseminated through a communications link to NOAA's HazCollect system. This allows distribution through NOAA All-hazards Radio and the Emergency Alert System (EAS) in cooperation with the Department of Homeland Security.
Although CAP is an open standard, there's no indication that the format has been picked up by emergency services outside of the US. That's unfortunate; this is exactly the kind of technology that should be consistent globally.
(Thanks, Art Botterell)
I believe that CAP will get international recognition and acceptance, but it will take time. The flexibility of it will help to have a long-live and the chance to be adopted gradually.
To have some example ... the EU project MEDSI (see www.medsi.org) incorporates CAP as one of the key vehicles to disseminate messages from and to a decision support system for crisis management and critical infrastructure protection...
Could this be used to create a "disease reporting system" (aka epidemiological surveillance or monitoring system) based on community observers or on patients themselves?
Imagine a blog where the blogger updates her health situation regarding (preferably non-pandemic) flu. Or, if you want to protect privacy, then report on one or more "proxies" - a teenager might tell about the current flu-state (yes or no) of her grandparents.
An agregator could provide estimates that would work like a "sentinel network" - not a physician sentinel network, but a patient (or proxy) sentinel network.
Negative information ("no, my grandma does not have symptoms of flu at the moment") would be useful too. Cases / (cases + non-cases).
CAP is as redundant as the ARC was. It's a rebranding of the same things that have existed for some time - the question is whether it will ever be used - which is the same problem that ARC had.
It's like calling collective intelligence, 'smart mobs'. ;-)
Does anyone know where you can get some of the CAP messages from the USGS?