During the first week of March, the coronavirus story shifted up a gear in Sweden. As people were returning from half-term ski holidays in the Alps, the rate of new confirmed cases — and the amount of related information from national and regional health authorities — jumped.

In the newsroom at Aftonbladet, Sweden’s top-breaking news site, the effort to keep up with developments intensified.

The task of manually staying on top of updates from around the country was time-consuming and required reporter effort that would be better utilised elsewhere. There was an idea that perhaps all the incoming relevant information could be gathered into a single feed, assisting with both speed and accuracy.

On the morning of March 5, the Aftonbladet news desk asked United Robots’ development team for help. After lunch, the first iteration of Coronakollen (Corona Watch) was up and running, sending alerts to the newsroom.

Alerts on new cases

Corona Watch looks for relevant new information based on criteria defined by the newsroom and then sends it to the desk. According to Aftonbladet managing editor Michael Poromaa, this solved the problem of monitoring the 21 regional healthcare authority Web sites every second, which would have been required to be the first to report on new corona cases.

“There was an immediate effect,” Poromaa said. “From having been, at best, second to publish new cases, we’re now more often the first, as happened with 60 new cases in Stockholm on March 10, for example. It’s extremely valuable for us that reporters get notifications through the robot alert channel in Slack.”

In its first week of use (March 5-12), Corona Watch generated 158 notifications. This turned into numerous updates on Aftonbladet’s live Corona feed (and a number of subsequent articles). The number of sources increased from 4 to 20.

News desk automation

There’s a lot of discussion in the industry about newsroom automation, and this can apply to various different processes. Using robots to automatically generate articles from large data sets is one application. A number of successful cases internationally — from The Washington Post to local media groups in Scandinavia — were recently covered in this recent INMA webinar.

While deploying automated content requires a clear strategy and reasonably sophisticated execution to really provide journalistic and business results, automating news desk functions is quick and generates immediate effects.

As with content automation, news desk automation frees up reporter time. It also means the newsroom can lean on the news desk’s tech to support immediate publication based on new developments. Nothing gets missed and updates are instantaneous. And, as the story evolves, new sources can be added.

United Robots works daily with Aftonbladet and other large and small media groups, providing news sites with automatically generated texts on topics such as sports, property sales, and traffic.

“In a way news desk automation is even more critical for small newsrooms with limited resources,” said United Robots’ CEO Sören Karlsson. 

The last couple of weeks are proof that where fast-evolving, very big stories are involved, automating how the news desk finds its information can support both the quality and the quantity of the journalism.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .