Met Office casts the API line

Interview: Chief digital officer Rich Carne talks about the strategy for meeting the public service remit through using application programme interfaces

It was a downturn in website traffic that prompted Rich Carne and his Met Office colleagues to think seriously about a strategy for application programme interfaces (APIs).

Rich_CarneSix years ago the organisation developed a mobile app that initially attracted a fraction of the 400 million visits per year to the website. But over the past three years there has been a crossover, so the app now gets about 400 million visits against the website’s 200 million.

“We got thinking about what was going to be the more important channel,” says Carne, chief digital officer at the Met Office. “We have been placing more investment into the mobile app, but the next step involves the emphasis on APIs.”

Carne has spent much of his seven years at the organisation focused on web applications, having led the effort to develop them for commercial and government customers in his previous role of head of IT applications. He now has a wider remit around digital and technology transformation, and is devoting a lot of time to an API management strategy.

Competing requirements

In doing so he has to balance two requirements that could come into competition. One is to fulfil the public service remit of the Met Office and make information on weather and climate change available as widely as possible. The other is that it is a government trading fund, with a responsibility to pay for itself through selling services – and that means data.

The strategy is aimed at striking the right balance, partnering with other organisations to push out data through their digital channels, but also developing specialised, high end data that can attract revenue in lucrative markets.  

It got under way two years ago and is still evolving, but has been grounded in three core principles: controlling access to the data; automating the monitoring of usage and regulating it when necessary; and building a community of developers.

Some of this involves working with major companies, most notably Amazon, on new channels for the data. Carne says this has included TV, Amazon’s tablet app, and its Alexa digital personal  assistant. The latter includes a back end API and voice recognition capability that makes it possible to provide contextual information about the weather, drawing on what it knows about the individual.

There has also been some work with Google.

Integration growth

“There’s a huge growth in platform integration,” Carne says. “Developers are using information from different services through APIs and aggregating to add value to their own business propositions. We’re responding to the way the market is going and trying to position ourselves within the market.”

It also involves giving individual developers and small companies the opportunity to use their data in their own apps.

“We want to see an increase in the number of people consuming our information via APIs into their own platforms. We expect some attribution, maybe our logo and at least our name, as we are a trading fund and have to consider reducing the cost to the taxpayer.

“If we can offset some of this by selling to the higher end consumers we will do that. I see the commercial bit as a means of supporting our public service message.”

A couple of the cornerstones of the strategy were already in place. DataPoint has been around for five years, a website that gives app developers a route to freely available data feeds.

It provides information on features such as radar and cloud cover for the UK, site specific data on temperature and winds, and weather forecasts for regions by text. All the developers need is an API key, the code that identifies the program and developer to the source.

In August of this year it had 14,000 registered users of whom more than 9,000 were using it daily, and the volume of data requests was about 25 million per month.

Carne says there is some scope to commercialise elements of DataPoint by making distinctions in the range of data and the quality of service.

WOW factor

The Weather Observation Website (WOW) almost reverses the processes, with the Met Office aggregating data from a range of private sources. It enables anyone to input their own observations, including photos, to provide a detailed picture of the weather.

A third channel, Beta Data Services, is under development. It is currently a closed environment as the developers and a group of partners experiment with its potential for exploiting the next generation of internet data.

Carne says the aim is to make it possible for developers to take specific sections from a massive block of several million data points to fulfil a specific purpose, and cites the example of the area around a flight path in aviation.

“We have users who want information on a specific part of the data cube, and we are developing open standards for developers to find relevant parts of the cube intelligently,” he says.

“We feed these through a web processing service that does the heavy lifting, passing on a few data points rather than monolithic terabyte files. Now we can map an overlay on the web map service and provide access to xml files.”

He says this will be aimed at high end customers, possibly including government agencies, with the potential to bring in revenue that can provide the balance against the spread of more basic data through the APIs. The potential is reinforced by the fact that weather data is very transient and, when it affects big business, has a high value.

Extending reach

While the development of Beta Data Services goes on, the work on bringing more developers into the fold continues. The Met Office has support engineers in place, has run hackathons with different groups, and is telling SMEs that they can use its data to create value added applications that support their business, and at the same time extend the organisation’s reach.

“We want more developers to use it because that will grow our reach,” Carne says. “If we work with a company like Google to develop weather warnings we immediately get the benefit of their massive reach; and the aggregators win because they get the rich, compelling content that drives traffic to their sites.”