Everything connected to everything else? Not in the way you think!

And though one says that one is part of everything,

There is a conflict, there is a resistance involved;
And being part is an exertion that declines:
One feels the life of that which gives life as it is.

Wallace Stevens, “The Course of a Particular”

–If there were ever a term in need of greater differentiation, granularity and detail, it is “interconnected” (as in interconnected critical infrastructures). Why?

Economists, engineers and system modelers often conceptualize interconnected critical infrastructure systems along the lines that Garret Hardin did 50 years ago for the Tragedy of the Commons: Imagine an interconnected critical infrastructure open to all manner of vulnerability and complex interconnectivity. That is precisely what you cannot assume on the ground. Empirically, there are different types of interconnectivity and these differences matter, something Hardin got terribly wrong when it came to livestock rangelands.

–By way of example, our research on a Vessel Traffic Service (VTS) of the US Coast Guard (USCG) found at least five major kinds of “interconnected” at work having sharp differences in the VTS’s real-time operations:

• Interoperability: Like the textbook interoperable energy utility (where electricity is crucial for the natural gas operations and vice versa), the VTS manages both vessel traffic and the regulated waterways that the vessels use (where managing the water ways affects management of the vessels and vice versa);

• Shared control variables: Water flows are a major control variable not just for VTS navigation purposes, but also for other infrastructures (most notably large water supplies and hydropower systems). This means that unexpected changes in how one infrastructure manages water flows can affect the management of the water flows by the other infrastructures (indeed, inter-infrastructural coordination around shared control variables was reported to us);

• Combined cycle of infrastructure operations: The USCG has a range of missions and operations, two of which are the VTS and the SAR (Search and Rescue) units. VTS combines with SAR to represent stages of this infrastructure’s operational cycle—normal operations and disrupted operations (VTS) along with failure and recovery (SAR). Not only are normal operations of the VTS already inter-infrastructural (by virtue of the shared control variables), but also the USCG’s Command and Control mission, including that for SAR, has an incident command facility and function for inter-infrastructural coordination during system failure and recovery;

• Variety of real-time configurations of interconnectivity: The VTS manages by virtue of resorting to a variety of interconnections with the vessels concerned. When VTS management of a common pool resource (the waterways) on behalf of inter-related users is disrupted or fails (e.g., because of defect in VTS communications), the interconnection configuration defaults over to the reciprocal one of vessel-to-vessel communication; and

• Inter-organizational linkages: USCG operations, including a VTS, are not only linked with other infrastructures through reliance on the Global Positioning System (GPS), but the Coast Guard’s position within the Department of Homeland Security makes it strategically located with respect to focusing on GPS vulnerabilities and strengths when it comes to the nation’s cyber-infrastructure.

Why do such differences matter? One answer: Once different interconnectivities are taken as the serious, really-existing starting point—”the life of that which gives life as it is”—we understand better how major approaches to risk management of critical infrastructures can be mis-specified and misleading.

–One case in point will have to suffice. What could seem more reasonable than a focus on system chokepoints and the most obvious way to do that is by focusing on where major infrastructures intersect or lie adjacent to each other on the ground, right? Wrong.

It’s wiser is to focus on how spatially adjacent or collocated structures and facilities are actually managed within their respective infrastructure systems. It is possible that a system’s chokepoint may be elsewhere than at the site of collocated facilities, and that the element collocated could be lost without its respective system flipping into failure.

Just because elements from two or more infrastructures are spatially adjacent does not mean automatically mean those infrastructures have “to coordinate” unless, say, shared control variables are involved or interoperability challenged.