Opening up public sector data will give value back to the citizen

There’s a push to open up public sector information stores, with a 2011 study saying the overall economic gains will be around €40 billion a year for the EU’s 28 members. To realise this value, public sector custodians must overcome their reluctance to give up control and learn how to share and use this information resource, says Pól Mac Aonghusa of IBM.

Q: What is the appeal of smart, open public sector data?

A: For open data I think it’s an interesting journey from a very centrally-held and closed data regime in the public sector, to a greater opening up. I think it’s healthy. Judging by peoples’ activities – the downloading of apps and the start-up companies that use open data – there’s a demand for it. People want meaningful data though – not data dumping or noise.

Q: Does the public sector in general have the capacity to use smart data?

A: For the public sector, the mindset is important. There’s no real tradition or motivation for publishing reams of information. Already there’s a barrier: public officials need to understand it’s a good thing to publish data. When they do it, they must prepare for success. Public authorities should ask themselves – what if people demand even more data and what happens when they don’t just want data on traffic cameras – for instance – but for a whole host of other areas too? How do you gear up?

The public sector needs to anticipate start-up companies relying on the data. If [this] is too overwhelming, what about exploring a public-private partnership to help manage it? If you’re thinking of having real active engagement through open data, you need to prepare for success.  

Q: Can you give a good example of open data in practice? 

A: We ran [a] traffic management pilot in Dublin – a city, like the majority, that had no way of visualising data globally. The city’s traffic control room looks after around 1.4 million commuters. They managed it in a very analogue fashion – using cameras on junction boxes and on traffic lights. Finding out the source of congestion, if this problem presented itself, was slow and involved checking some 40 television screens.

What we did to change this was to take global positioning system data from Bus Éireann, the city’s main bus and coach operator, and use the buses as a proxy, or sample, for traffic throughout the whole city. From this, we built a live map of traffic.

Q: While a live map can give greater information on traffic flows, how does it go towards explaining the causes of traffic delays?

A: We built an artificial intelligence engine. We fed it a map of the city. We gave it data on traffic from as far back as we had it, which is about a year’s worth of data. From that, the machine learned about traffic behaviour. It built a picture of what traffic in Dublin is like on an average wet Monday morning, for instance.

The system is able to formulate three hypotheses for why traffic is moving slowly based on information it can dynamically pull from the internet. It will go on events websites and will then know if there’s a football match or a Madonna concert that’s holding up traffic. It’ll also suck in Twitter feeds to see if anyone’s talking about a road accident on road X and look at city council maintenance databases to tell us if anyone’s digging up a road.

Q: How do the traffic controllers interact with this enhanced information?

A: Streets are designated red, green and amber – red being bad, green being fine and amber being something maybe you should worry about. This information puts them ahead of the curve. Instead of trying to look at 40 television screens, they now can see one screen that can tell them, “I’m 90 per cent certain it’s a water mains problem: you should call the maintenance people.”

The same system can predict traffic two hours into the future now. That’s forever in traffic management terms. With reasonable confidence we can now say, all else being equal, here’s what will happen during rush hour. We’ve tested this for a year and we’re pretty confident.

Q: How does this example translate to other sectors?

A: It caught our attention because when we started talking about the concept of a smarter planet, we talked about connecting things together more effectively – we’ve just seen how it can work in our traffic example above. If we look at healthcare, for example, we see that we tend to deliver it in chunks. If you’re sick, you go to the doctor. He or she doesn’t talk to your school teacher or your social worker. We started thinking about joining up all of these touch points, with particular focus on the high-risk, high-need population.

In an integrated care system, what if these people could share information? They could sit down together and consider each case as it comes before them. If a patient comes with diabetes, we’ll look at the best treatment and ask ourselves what this could be, taking a broader range of factors into consideration. Maybe it’ll be a weight-loss programme – because there’s a correlation between diabetes and weight. Under the integrated care model, you’d only consider a weight loss programme if you knew the patient could keep to the programme. Do they have the support network at home, are they motivated, can they exercise? The different care providers, together, could adjudicate on this.

Q: Are you seeing any pushbacks to an approach that uses open data?

A: For something like traffic management, no. For healthcare, of course it’s inherently a sensitive space. We need a fail-safe so people only see the level of information they’re entitled to see. A doctor should be entitled to see more medical data than a policeman: that’s a fair statement to make, anyone would agree.

Data commissioners are involved a lot – and this should remain the case – to adjudicate on the 50/50 calls. We have regular conversations about anonymising x, y and z. It’s a very interesting space to be in though. When you’re in research, one of the complaints is, “I’m doing all these maths but I don’t see the impact.” Will anybody care, will life be better and will anything change? Here especially we can see the potential to positively impact on peoples’ lives.