The volcano and the virtual

What’s infrastructure?

Colleagues at the Berkman Center have spent a lot of intellectual energy on this question lately. We’ve been lucky enough to have Christian Sandvig, communications scholar and technology critic, with us this past year, and he’s convened a group dedicated to the discussion, dissection and understanding of the infrastructures that make our contemporary world possible.

I, for reasons of a transportation infrastructure that makes it expensive, environmentally irresponsible and inconvenient to commute the 300 miles round trip to Boston more than a couple of times a month, haven’t been part of Christian’s infrastructure group. And so I don’t know whether my definition of infrastructure is wrong, derivative or merely flip:

Infrastructure is the stuff we ignore until it breaks. Then it’s the stuff we’re stunned to discover we’re dependent on.

I wrote about this idea fifteen months ago as I tried to figure out an apparent paradox – the very existence of some infrastructures make us feel more connected to the rest of the globe, even if we seldom use that infrastructure to make non-local connections. I ended up suggesting that we start making maps of how traffic flows on top of infrastructures and contrast these to the maps of the infrastructure themselves, and that we pay very close attention when infrastructures break, because we learn more from dysfunction and failure than from their transparent success.

So the eruption of the Eyjafjallajökull volcano and the resultant closure of northern European airspace seems like a great time to think about air travel.

At least six Twitter friends were stranded in different corners of Europe on trips interrupted by volcanic ash. They’re a good-humored bunch, despite what’s obviously a frustrating situation. One friend took extraordinary measures (including a 24-hour bus ride) to get from one corner of the continent to the other to give a speech. Another did his best to enjoy an involuntary trebling in length of his Barcelona holiday. All were, I think, reassured by the social permission that comes from a natural disaster – it’s hard to be angry with someone for missing a professional commitment when they’re blocked from travel by the mighty god Vulcan. It’s starting to sound a little bit like a hemispheric snow day – perhaps the best example of this was the TEDxVolcano conference, organized in less than 24 hours to take advantage of the fact that a lot of very smart people were stranded between gigs in London and willing to get together and live-stream a conversation.

The immediate consequences of airspace closure are clear. People were stranded on the wrong continent for quite some time. Perishable commodities – fresh flowers, produce, some pharma – were in short supply and their producers lost revenue. Airlines lost revenue, and will probably lose more even after the disaster, as nightmare stories of being stuck in Europe or North America tend to interfere with summer vacation fantasies.

As Jeff Jarvis noted, airlines aren’t really networks. They have a lot of trouble reconfiguring to cope with partial system failure. It would be great if everyone could have hopped on the train from the Netherlands, Barcelona and Berlin and converged in Rome, where flights could run around the clock to ferry stranded travellers to the US. That’s how internet engineers would try to address the problem. My guess is that Rome’s airspace is already pretty congested and can’t double or triple the number of flights that land at FCO… while most internet infrastructure is built to sustain peaks of activity that dwarf everyday traffic, which makes this sort of rerouting possible. And, unlike well-architected data networks, there’s not a ton of redundancy built into airline networks. Hundreds of smaller airlines use Heathrow as their main hub for international traffic. Qatar wasn’t directly affected by the ash cloud, but much of their travel hubs through London and the disruption likely had some financial impacts.

What’s interesting to me is what will happen if the airspace closure wasn’t a one-time thing. As many commentators have observed, it will be bad for airlines, good for trains, ocean liners and videoconferencing companies. But I wonder whether the changes won’t be more subtle and pervasive.

Because the infrastructure of international air travel is generally so reliable, we assume that physical presence over long distances is constrained primarily by time and money. A conference organizer in London invites me to speak at an event – she’s conscious that I’ll say no unless she pays for the airfare, and usually works to make the request for my time as modest as possible. Because we both assume that air travel almost always works, we can agree on a plan that would have seemed absolutely ludicrious even a generation ago – I’ll fly in Thursday night, speak Friday morning and head back early on Saturday, spending less than 24 hours on the ground. (Yes, I’m scheduled to do this in June. Yes, I understand how ludicrious this is.)

Visualization of CO2 impacts of volcanic eruption, aerospace closure from Information Is Beautiful.

It’s possible that Eyjafjallajökull could change this. If a 24 hour trip to London has a significant risk of becoming a 5 day trip to London, the calculus changes. As much as frequent travellers gripe about delays and cancellations, they’re pretty infrequent, and mass delays like the ones currently being experienced are downright rare. If they become commonplace, I personally would expect to say no to travel lots more often and do a lot more appearances via Skype and videoconferencing.

When the price of gas shot past $4 a gallon in the US, people started selling off their large SUVs and buying hybrids. Almost every organization I work with has a policy that they’re going to start travelling less, using more telepresence and trying to minimize their environmental impact. But I haven’t seen those policies actually change behaviors – I still get asked to travel far more than I get asked to appear via Skype video. Getting stranded, and the threat of possible future stranding could be an impetus towards actually changing our behavior as regards how we hold meetings, conferences and other events. And perhaps we’d actually get better at doing virtual events.

I gave a talk at MIT earlier this month that reminded me both how powerful and how limiting videoconferencing is. I moderated a panel in an auditorium at MIT with a live audience, but all my guests appeared over Skype. When we initially set up the room, there was no way for me to see the monitor where our guest speakers appeared. I found that so disconcerting that I ended up moving the monitor and repositioning my chair, so I could make “eye contact” with the panelists. They, however, couldn’t see me… though they could see Chris Csikzentmihayli, who was managing the feeds. More than one of the participants sent a note expressing their happiness with the event, but commenting on the challenge of not being able to see a moderator. Obviously, we’ll do better next time, putting an iSight camera on my face and perhaps another on the audience.

But the mechanics of the panel aren’t the problem – it’s the social contact that falls off in virtual events. As much fun as my conversations with panelists were, the unexpected, serendipitous connections were the ones I made with a Nigerian science journalist and an Iranian hacker in the bar after the panel. I suspect that tools like Twitter could make these virtual events a lot more social than they’ve been in the past, but it’s very hard to imagine getting the same level of informal interaction that we have in real life

Or maybe we haven’t tried hard enough. If travel between North America and Europe becomes difficult or impossible – i.e., if the infrastructure we usually ignore forces us to pay attention to the challenges of physically convening meetings – maybe we’ll force ourselves to get better at being virtual. I put this question to a group of friends who’ve organized a number of conferences and got some good suggestion: perhaps we could set up terminals running Skype in the venue, so that the audience could crowd around and ask speakers questions after the formal presentation? Perhaps we combine this phase of the event with a cocktail hour, where the speakers are encouraged to open a beer and continue the conversation with whoever comes by to talk to the screen. (One friend insists that there’s nothing more depressing than drinking alone and suggests that whoever presents remotely assemble her own audience, even if it’s an audience of one, to root her on and celebrate alongside.)

What happens when infrastructure’s no longer reliable? If it’s no longer invisible, do we start questioning our dependence on it? Is that what we need to motivate us to look for different solutions?