“So what are the most important developments in journo tech right now? Which tools and concepts will help us create really interesting content? And what about improving production workflows?” That’s a set of questions we get quite often. Providing definitive answers is impossible. What we can do, however, is share a couple of insights from our work. Here’s a list of tech that struck us as particularly interesting, promising, or useful this year.



Illustration: Clker-Free-Vector-Images

5G is the fifth and latest generation of mobile communications standards. It’s a “bigger, better, faster, more!” thing, and it certainly won’t go away. 5G will feature 10x the bandwidth of the current standard, allowing for the streaming of sophisticated content (HD video, VR, etc.). It will also reduce latency, and boost the number of local clients, thus improving wireless internet of things (IoT) infrastructure, e.g. in a studio environment with all sorts of gadgets and workflows. On the negative side, 5G technology may interfere with “traditional” C-band satellite transmissions. DW will continue to watch the new standard and experiment. Large scale 5G rollout in Europe isn’t expected to begin before 2020 and will probably take a couple of years.

Apache Kafka

apache kafka

Illustration: Apache

Apache Kafka is a powerful open source software that can be used for message brokering, website activity tracking, log aggregation, or big data stream processing. It has been around for seven years now, but mainstream success in the media business is relatively new. Kafka serves as a “circular buffer” and basically helps platforms and services handle a huge number of requests in a rapid and reliable way. A very descriptive example is the use of Kafka by Netflix: The US media services provider has implemented the system to make sure binge-watchers around the world can request, play, pause, and resume a vast number of videos at the same time, without any (major) glitches. Legacy broadcasters and media organizations have been following suit.

AR Journalism

ar journalism

Illustration: Google

AR is short for augmented reality, which is not exactly a new thing, but currently making a huge comeback — especially due to powerful SDKs like ARKit (Apple/iOS) and ARCore (Google/Android). The latest generation of AR is great for journalistic storytelling, as producers can create captivating quality content in a matter of days (or hours). Unlike VR, AR stories/experiences provide information “in situation”, by projecting extra layers on top of current surroundings (via gadget screens or glasses), and thus achieve immersion and interaction without overwhelming or isolating the user. Our team has been looking into “modern” AR for almost a year now, creating prototypes at the DW Lab or at AR hackathons, and putting together a couple of media innovation proposals.

Distributed ledgers

dl blockchain

Illustration: mmi9

Distributed ledgers are – of course – the umbrella term for blockchain technology and all that jazz. We just thought we’d avoid the hype and be a little more level headed here; especially since 2018 saw a lot of people jump on the crypto bandwagon. However, while some of the use cases out there are questionable, the tech itself (which is a decade old, by the way), is interesting and potentially sustainable: A properly designed blockchain can ensure that a lot of people who don’t know and/or don’t trust each other can do fast, reliable, scalable business via a private or public record, which is constantly peer-reviewed (by machines), and virtually unalterable. It can also save everybody thousands in unnecessary notary and lawyer fees. Journalists and the media industry can benefit from blockchain tech via new models for content licensing (e.g. smart, transparent ledgers for texts, pictures, audio files, videos etc.). Our H2020 project Bloomen is currently looking into this.



Illustration: ahmedgad

HLT is short for “human language technology” and comprises a lot of tools and services that are based on the latest leap in artificial intelligence (AI), or more specifically: the power of bleeding-edge artificial neural networks. Common HLT “features” are ASR/STT (Automatic Speech Recognition/Speech-to-Text), NMT (neural machine translation), and TTS (text-to-speech). Lesser known outside of HLT circles, but equally exciting: algorithms for automatic summarization or topic clustering. Consumers/audiences mostly encounter HLT via voice assistants (in smartphone and on kitchen tables), journalists and media producers have started using it to create multilingual versions of their content: For example, HLT services allow for the automatic transcription and translation of interviews, which can then easily be furnished with foreign language subtitles or voice-overs. As of December 2018, DW Innovation is involved in two major HLT efforts: SUMMA (for media monitoring in a lingua franca) and news.bridge (for multilingual content creation). Follow-up projects are scheduled for 2019 and beyond.



Illustration: Bokeh

Jupyter refers to Jupyter notebooks — the product of a powerful open-source application of the same name that has been picking up steam for a while now. Just like the Greek god (who is spelled with an “i”, of course), Jupyter notebooks are powerful and feature-rich. Users may combine text and photos and videos, and — most importantly — code in all sorts of languages (e.g. Julia, Python, and R). They may also execute, test, and alter that code within the notebook, a truly outstanding feature. Naturally, everything can be accessed locally or in the cloud, shared in different ways, and collaboratively expanded. Especially when combined with container technology like Docker or development platforms like Github, Jupyter notebooks become state-of-the-art digital binders, that can be very useful for interdisciplinary teams with a knack for data science. DW has been using Jupyter to create, test, and document exclusive DDJ stories (data-driven journalism) since late 2017.



Illustration: GDJ

OBA is maybe the most cryptic abbreviation in this list, representing a set of extremely relevant, yet still rather unknown technologies. Object-based audio, also referred to as next generation audio (NGA), might very well trigger a radio renaissance in the coming decade. The basic concept: When recording audio, use a huge number of sources/objects, enhance them with metadata, and later on fire up a “renderer” to play back what you’ve captured in a tailor-made fashion. The result: High-end multi track 3D sound for young podcast aficionados, but an old school stereo mix with an extra loud narrator voice for their grandmothers. OBA also allows for other forms of personalization, e.g. multilingual features, outtakes, and extra in-depth content. For media producers, OBA ultimately means simplified and haromized workflows: There’s only one production, and it covers all features, target formats, services, and platforms. At DW, we’ve only just begun studying OBA, e.g. by following the (recently) completed H2020 project Orpheus, or by experimenting with 3D sound at our lab.

More interesting journo tech

“Anything else? What about all the stuff they mention over at Nieman Labs and Wired Magazine? What about other DW projects?” Well, there’s always more, but we’re primarily a team of people experimenting with journo tech – not a team of people reporting on it 24/7. Hence the personal list of highlights. As for our other projects and products, there’s a strong general focus on AI, which is used to verify content and fight misinformation (Truly Media, InViD, WeVerify), but also to come up with a more personalized news diet (CPN). DDJ and data science are a cornerstone of EU BusinessGraph, while the MultiDrone consortium is figuring out how to best use those stupendous flying gadgets for journalistic purposes. Last, but not least, we still have an eye on virtual reality (VR), e.g. in our our H2020 effort V4Design. Things won’t get boring anytime soon.

(Alexander Plaum)