Exactly 10 years ago, I worked on a project at the European Space Agency (ESA) exploring who, if anyone, would actually use the data from Europe’s Copernicus satellite system – built at a cost of nearly €7 billion to European taxpayers – and what they might build with it.
The core idea was that European institutions like ESA could invest in the technological infrastructure of expensive satellites and sensors, and that the private sector would eventually build applications on top of it far faster and more cheaply than any government could.
A decade later, I found out just last month that a man in Enniscorthy has done exactly that. In two months, motivated by the devastation that Storm Chandra inflicted on a country that still has no flood-warning system, Gavyn Pedley built his own by fusing real-time data from Met Éireann, the Office of Public Works’ (OPW) river gauges, Teagasc soil moisture readings, and Europe’s Copernicus satellites.
Back in January, when an Oireachtas committee asked the OPW why the Republic still had no local flood-warning system, the answer was that it would take five to 10 more years to create one, that the first stage alone had taken seven years instead of five, and the next stage of localised alerts sent to people’s phones had no budget and no timeline.
READ MORE
A credible question at this stage is how it could take the State 10 years to build what a local data consultant has apparently done in two months; the question itself is rage-inducing and symptomatic of the lack of State capacity, a theme woven through recent political narratives.
I sympathise with the Government and acknowledge that the gap between two months and 10 years to create a flood-warning system is the difference between building a piece of technology and building the institutional architecture that determines what happens when that technology tells you a flood is coming.
What Pedley built is a prediction model – one that may or may not prove reliable at scale – that tells you the river is rising and your street is at risk. Its existence demonstrates this is now a technology problem that a single person can solve with publicly available data and a laptop.
But his system could, in theory, send a perfect alert to every phone in Enniscorthy tomorrow morning and it would change nothing because what the State has failed to build is everything that would need to happen after that prediction is made. Such as which of the multiple departments that share responsibility for flooding would receive the alert, who would be liable if the warning turned out to be wrong, and who would deploy sandbags, at what trigger level and under whose authority. These are questions of institutional design and require that the mandates, protocols and legal authorities are able to convert information from an app into a co-ordinated public response.
What would take the Republic a decade or more to build, and starting from a position of zero, is co-ordination across numerous government departments, Met Éireann, the OPW, 31 local authorities, and the Office of Emergency Planning – agencies that co-operate in the aftermath of each emergency but have never been required to design a standing system together in advance. I understand why this is taking a long time, because getting agencies to build something proactive and permanent is genuinely among the hardest things a government can do, even for the most capable.
But the difficulty of a flood-warning system is precisely the reason that co-ordination has to be treated as the thing that must be built, rather than accepted as a permanent absence that the State simply works around; the status quo for decades.
Without that co-ordination, the State retains no institutional memory, as the lessons from each failure remain trapped inside whichever department or agency happened to handle it, never reaching the many other disjointed ones that will handle the next crisis.
The flood-warning system is not the only place where this absence is reproducing itself. The Critical Infrastructure Bill was introduced in the same month that the Galway Ring Road was approved – a road that will kill the case for the Galway Luas, a far superior solution to the city’s transport problems, condemning the city to car-dependent sprawl. The proposed motorway will effectively repeat every mistake Dublin made with the M50, and the disbelief at a country repeating mistakes it is still paying for is the same disbelief as the 10-year flood-warning system. These are, after all, the same institutional problem.
[ Flooding in Enniscorthy: Locals sceptical of offical responseOpens in new window ]
There is a precedent for getting this right and it comes from the same space industry whose data Pedley used to build his flood predictor. When Nasa and the US government – including the Pentagon – decided to develop the next generation of space launch vehicles, they made what turned out to be the most consequential decision of the programme. This was that the government would choose not to build the rocket, and would instead do the hard work of defining the requirements and constructing the project oversight that allowed a private company to build it. Enter SpaceX, the extraordinary success of this model.
In this way, the government maintained co-ordination across the whole system while allowing the private sector to move quickly and cheaply on top, in a complex interplay that leveraged the best of both worlds.
That decision, though, required clarity about the government’s purpose, and a deep understanding of what it existed to do, as well as what it could confidently leave to others. It is precisely this clarity that the State lacks, because nobody has ever articulated what the State’s job actually is. This means the Government oscillates between attempting to build things itself with decades-long delays and outsourcing them in short-term bursts to consultants and contractors. As such, it never establishes a strategic framework for which activities it is best suited to build and which it should release, all without retaining the institutional memory from either approach.
The question is whether the State will build the one thing that only a state can build, which is procedural capacity that allows it to know what it should do and what it should leave to others. Eventually, this could mean that when someone like Pedley builds a much-needed technology in two months, the State is ready to use it.
