Machine learning is still in its early stages at the Βι¶ΉΤΌΕΔ. While we've been working on it for some time within Βι¶ΉΤΌΕΔ Research & Development, we have only more recently started exploring how we can use these tools in our production areas.

Βι¶ΉΤΌΕΔ iPlayer uses machine learning to provide personalised recommendations to audiences based on their viewing habits and detects when a programme's end credits start, as it's been found that more people respond to recommendations when they are shown at this time.  

Overall, recommendations on Βι¶ΉΤΌΕΔ iPlayer have increased the time audiences spend with us.

Elsewhere, recommendations are also being used successfully by Βι¶ΉΤΌΕΔ News, and our metrics are showing us that this is driving our audiences to look at more pages.

This barely scratches the surface of the potential of machine learning, though, and at the moment we can't even answer basic questions such as how much content we have on a particular topic.

This is because we have a lot of different systems that need to be searched through and because our approach to classifying or tagging content hasn't always been consistent across different parts of the Βι¶ΉΤΌΕΔ. 

So for example, the topic of ‘pirates’ means something quite different for children's content than it does for news, so how can we differentiate? 

This is where machine learning can help. We are currently exploring whether automatic tagging is possible, to improve consistency. This will help us surface more relevant content to audiences and bring the right content in front of the right user, at the right time.

The Research & Development team has built a number of tools to help with this. One tool searches in articles for keywords, such as the names of people and places, and another extracts topics by finding words that are most likely to be linked to them. So, for example, it can tell that an article using the words revenue and share price is most likely a business article.

We now need to bring together some of these different initiatives. To address this Anne Bulford, Deputy Director General, and Matthew Postgate, Chief Technology and Product Officer, have sponsored the Connected Data programme.

Connected Data is building a shared data science and machine learning platform that will make data available about all Βι¶ΉΤΌΕΔ content, and provide tools and services that enable product teams to use, enhance and extend the platform to deliver richer audience experiences, more easily.

We are also working on creating training materials to educate developers and colleagues more widely about machine learning. A Βι¶ΉΤΌΕΔ Academy website with resources has been developed and we are running courses.

We are very much at the beginning of our journey with machine learning, but it could really have a huge impact, not just for our audiences, but also for everyone at the Βι¶ΉΤΌΕΔ.

No, not a rival to ITV’s show.

But an update on what we’re doing with voice devices.

Last year,  and in December , our Amazon skill.

On Christmas Day, listeners across the UK could find our live radio stations and our podcasts easily when they unwrapped and started trying out their new presents.

Here’s some first numbers on how it’s going.

• We’ve seen 1 million unique browsers use our voice app
• Radios 1, 2 and 4 have so far proved to be the most popular stations when it comes to requests
• People seem to love listening to the radio on voice devices - they’ve been slower to ask for on-demand content

I’m guessing that will change as people get used to interacting with their devices.

This seems to be on course to be something lots of us do. Almost two thirds of UK adults claim to own a device with voice (including mobiles) with one in ten claiming to have access to a smart speaker - this has more than doubled in the past year. This is faster than the uptake of tablets, at the same stage.

So what next? We want to discover new types of content, content that’s been made for Voice, stuff that’s never been possible before. The platform is interactive. It’s not constrained by the structures and formats of linear schedules. As with all our digital experimentation, we’ll try different things, some will work, some will need improving.

One of the first areas we’re experimenting with is programmes for children.

Last year  to make sure we can reinvent how we serve these audiences, building on what we’ve already done with iPlayer Kids

So with voice we’re going to start making Βι¶ΉΤΌΕΔ audio content available to kids.

Later this spring, we’ll be creating interactive experiences around some of our most familiar strands for young children.

We'll start small but add new experiences and narratives as time goes on, as we learn about the kinds of interactions that children really like and benefit from. We also aim to offer some of our best children’s stories for bedtime in audio form, but activated by voice - building ways for children to choose a story for themselves.

Children’s content is of course an area where trust matters, and we will make sure that the experience is safe for kids - it will be available through a dedicated skill for children.

We’re also thinking about other genres - such as news, current affairs, food and music.

We’ll blog about those soon too.

The review concludes that Βι¶ΉΤΌΕΔ R&D delivered significant value to the Βι¶ΉΤΌΕΔ, audiences, the creative community, the wider industry and the UK economy. Its cost-benefit analysis conservatively estimates that every pound spent by Βι¶ΉΤΌΕΔ R&D delivered a return of £5 - £9 to the UK.

This analysis takes a range of factors into consideration, including economic benefits and efficiencies delivered to the Βι¶ΉΤΌΕΔ, audiences benefitting from new or improved services, and other factors like patents and technology standards that benefit the industry and broader society.

As I say, it’s an exhaustive review so I want to provide a snapshot for anybody interested in learning more about our work in this post. If that whets your appetite, I encourage you to read the full report and there’s plenty more on our and too, where we publish regular updates on our projects.

 

What Βι¶ΉΤΌΕΔ R&D delivered for UK audiences and wider society

Our most immediate and high-profile contribution was to Βι¶ΉΤΌΕΔ iPlayer. We had helped the Βι¶ΉΤΌΕΔ take its first steps online in 1994 – soon after the invention of the World Wide Web – and continued to help set the Βι¶ΉΤΌΕΔ’s direction and strategy.   Our insights, including trials of new forms of content delivery like online streaming, were critical to iPlayer’s success after its launch in 2007.  The work has continued to help make Βι¶ΉΤΌΕΔ iPlayer more widely available, particularly integrating it into the wide range of TV platforms, which is now one of the most popular ways to enjoy programmes on Βι¶ΉΤΌΕΔ iPlayer.

Improving picture quality was a major focus and we helped ensure that free-to-air terrestrial TV viewers were not left behind by the move to HDTV. Βι¶ΉΤΌΕΔ R&D led an international committee developing a standard (DVB-T2) that was more efficient and resilient than its predecessor. We believe R&D’s contribution helped bring forward the launch of HD on digital terrestrial TV for licence fee payers by three years. This work continues with our Ultra HD trials and inventing a form of High Dynamic Range for the broadcast and streaming industries called Hybrid Log-Gamma.

We also provided significant contribution to Freeview Play, Freesat and YouView, working collaboratively with our industry partners. These platforms helped ensure public service and free-to-air UK broadcasting – whether broadcast or online – remained universally available, prominent and widely used.

Βι¶ΉΤΌΕΔ R&D was also involved in the Βι¶ΉΤΌΕΔ’s coverage of major events, using experimental technology to provide new experiences and enhanced coverage. For example, we partnered with NHK to showcase the opening ceremony of the 2012 Olympics in Super Hi-Vision, provided 3D binaural sound for the Proms, and created a smart video player that could learn, and then recommend, content to Glastonbury fans.

Elsewhere, we helped make Βι¶ΉΤΌΕΔ programmes more accessible through improved subtitles, and our Piero sports graphics system allowed programmes like Match of the Day and the Βι¶ΉΤΌΕΔ’s Olympic coverage to highlight extra detail on-screen. This meant audiences could see the speed of a ball, how closely a player was caught by the offside trap, and how Tom Daley’s splashes compared to his rivals in 2012 for the first time.

More recently, Βι¶ΉΤΌΕΔ R&D developed the early prototypes for the Βι¶ΉΤΌΕΔ micro:bit – a small programmable device that was given freely to all Year 7 or equivalent pupils across the UK. Our department also led the Βι¶ΉΤΌΕΔ’s first virtual reality experiences, helping us better understand the potential of this format to deliver compelling experiences to our viewers.

Collaboration with the creative community

Collaboration has always been critical to Βι¶ΉΤΌΕΔ R&D fulfilling its objectives, and this will be even more important in future. The Βι¶ΉΤΌΕΔ will face new and different market and operating environments, with fast-paced technology change and disruption, new types of competitors and different standards that drive and shape the global media ecosystem. Last Charter we began collaborating in new spaces with new partners and in new ways.

Through initiatives like Βι¶ΉΤΌΕΔ Taster, Connected Studio and Βι¶ΉΤΌΕΔ News Labs we worked with hundreds of new partners to develop ideas, concepts and technology for Βι¶ΉΤΌΕΔ online. Βι¶ΉΤΌΕΔ Taster, where we showcase many of these experiments, made 250 projects available for audiences to try. Connected Studio engaged with over 520 companies from the creative industries, resulting in 168 collaborative projects. And Βι¶ΉΤΌΕΔ News Labs worked with over 50 news organisations, universities and research bodies. This will continue to be an important theme underpinning much of our work in the next Charter.

Together with the creative community we explored how Object-Based Media might shape the future of content and production. Here, all of the individual parts of a programme – like individual shots, the presenter, the subtitles, the music or soundtrack, as well as data – can be reassembled by the audience, not just the production team. There are limitless possibilities to new experiences that we, and others, could create for audiences in an Object-Based world.

Some project examples include personalised video, which adapts scenes to a person’s personality and preferences; an interactive cookery show that changes in response to your decisions; and a weather forecast that can be adapted to a user’s needs and preferences – like their location and whether the presenter should use sign language or not.

Delivering value to the wider industry and the UK

Collaboration is also important to our work with the wider broadcast and technology industries. We have played, and continue to play, a major role in global industry forums that determine standards for the World Wide Web, 5G technology, digital television and video/audio compression amongst many others.   Our staff contributed to the development of numerous industry-wide technology solutions, including broadcast and online video compression, and the many standards needed to bring High Dynamic Range pictures to screens.  

We also set up a major partnership to unlock the potential of data in the media, bringing together academics, industry and colleagues from across the Βι¶ΉΤΌΕΔ. This long-term effort will help with the Βι¶ΉΤΌΕΔ’s mission to create a more personal Βι¶ΉΤΌΕΔ, and also help us better understand and take advantage of artificial intelligence and machine learning in broadcasting.

Other innovations start in Βι¶ΉΤΌΕΔ R&D and we then work with the industry to develop and ultimately adopt them. A good example is the Internet-based broadcasting system we pioneered. At the moment programmes are made using expensive and very specific production equipment, but in future the whole process could use fairly standard Internet technologies instead.

There are huge operational benefits to a system like this, such as being able to post-produce and mix programmes remotely from a web browser, meaning broadcasters could cover more events, more cost-effectively. And there are exciting creative benefits too, with the potential for more interactive, immersive and personalised content – like we have been exploring with Object-Based Media.

The Internet will continue to play an increasingly important role in the future of broadcasting, and our Internet-based approach is gaining adoption in the wider industry. A major focus over the next Charter will be to continue working closely with them to develop, test and standardise how content will be produced, stored, distributed and consumed in this new world.

We will need to manage the complex transition from traditional broadcast technology to Internet-based technology behind-the-scenes, while preserving the prominent, universal availability of public service content on-screen.

Making broadcasting more efficient and cost-effective

We led the initiative to move the broadcast industry from tape to digital file delivery. By successfully developing and establishing a single standard, we were able to dramatically reduce cost and complexity in the UK production industry. As a result, many global regions already have, or are looking to follow the UK’s lead.

We also developed the initial prototypes of the Βι¶ΉΤΌΕΔ’s in-house Content Delivery Network (CDN) - known as BIDI internally (Βι¶ΉΤΌΕΔ Internet Distribution Infrastructure) - which now carries a significant proportion of the media our audiences stream over the Internet. Having our own CDN - instead of relying solely on third-party networks - means we can monitor the performance of our streams more closely, make changes more quickly and easily, and manage costs more effectively. It also helps us manage the load the Βι¶ΉΤΌΕΔ places on the UK’s networks as streaming becomes more prevalent, as well as reducing the buffering for viewers.

I mentioned previously that our work on free-to-air HDTV made the technology more efficient. In practice this meant that - together with video compression technologies we also helped develop - we freed up enough broadcast capacity to initially launch three HD channels without having to reduce the number of standard definition channels. We continued to optimise these technologies and were later able to incrementally launch three additional HD and two SD channels.

As you will have gathered by now, there’s a lot in this report. We are immensely proud of what we have achieved.  But we are not complacent.  Some of the most significant advances in broadcast technology began in the Βι¶ΉΤΌΕΔ’s R&D department and we will continue to invest in R&D that best serves the interests of our audiences, the wider broadcasting industry and the UK’s creative economy, while always providing great value for money.  In practice, that means we are willing both to take risks on things that may not work and to take the decision to stop investing in things that we don’t believe are delivering value.  

There is so much more to come. What would a Blue Planet 4 look and feel like? What will it be like to talk to the Βι¶ΉΤΌΕΔ on smart speakers? What role can artificial intelligence or AI play in production? As this report shows, we share much of our research and insight publicly. If you want to follow our progress through the next decade as we try to answer those and other questions, you can find us on social media and as well as on our .

Andy Conroy, Controller, Βι¶ΉΤΌΕΔ Research and Development

• See the
• Follow

Over 70 days, from our anniversary on Thursday 29th September, we’ll be supporting both existing and new talent, from poetry to new music. For instance, in order to bring new music to more people we have teamed with Sound and Music to embed composer Matthew Kaner in our output - he'll be composing a new piece every week for the next 70 days, music that will be heard on Radio 3 Breakfast. The was premiered on Tuesday 26th September.

We’re also showcasing new poets in , a journey through the archives of our poetry output which sees greats like Sylvia Plath, Dylan Thomas and W H Auden reading their own work through to new works from poets like Alice Oswald, Simon Armitage and Liz Lochhead creating new poems for us. There's much more, as you'll see from some TV trails showcasing work we have commissioned which begin to be shown on Friday 30th September.

In tandem with our experimental content creation spirit, what many might not realise is our relentless commitment to quality sound. Classical music is so complex and wonderful in its sound makeup we need to translate that experience in a concert hall for audiences at home so they can feel immersed in the performance. Radio drama by the same token is such an intimate art form, that sound quality can really add to and enhance that experience. The Third programme was the first in this country to broadcast in stereo, Radio 3 experimented with quadrophonic, and to date we are proud of the HD 320 kps HD uncompressed sound we give listeners online, and we are looking to see if there are practical ways of taking this quality even further if this offers an advantage to the listener.

We have previously broadcast the Βι¶ΉΤΌΕΔ Proms in HD sound and quality surround sound, but this summer we’ve been undertaking some very exciting trials in binaural recordings. It’s the first time we have recorded the proms in binaural and the signs are good. Some early data back from the trials is encouraging, with over half of all who tried it giving it 5 stars – which means it was in the top 10% of all sound pilots to date and the people rating it was double the average for Βι¶ΉΤΌΕΔ sound trials. All of this means our audience love the technology. It shows our audience are hungry to have the best quality audio.

Music isn’t the only space we are experimenting, next week our evening of Beckett plays which will be broadcast as part of our 70th anniversary season on Sunday 2nd October featuring Ian Mckellan and Stephen Rea is one such broadcast. The plays will give an insight into the development of Beckett’s style and approach to sound, and it felt an appropriate thing to use truly immersive sound for, through binaural. Tom Parnell will give you a taster later. We’re also broadcasting in binaural as part of our anniversary, some J.G. Ballard mini dramas which air as Between Ballard’s Ears, the plays have been adapted by Frank Cottrell Boyce and Brian Sibley.

You’ll also see further developments in how we reach new audiences in new platforms, with more shareable content being developed that will present our output in different ways, using the Βι¶ΉΤΌΕΔ music app and other new technology in the pipeline. Then we are looking at how we use technology in different ways to enhance the experience of a live classical concert. So the Βι¶ΉΤΌΕΔ Philharmonic are launching a series of nine free concerts, something only the Βι¶ΉΤΌΕΔ could do with its fantastic performing groups, , which will present a different way of experiencing a concert. The concerts will also be broadcast on Radio 3.

Using ground-breaking new technology from Βι¶ΉΤΌΕΔ Research and Development, the concerts will create a more immersive musical experience for audiences both in the venue and across the country. The performance and interval discussion will feature in an enhanced live video stream online, which will enable the audience to delve closer into the orchestra, receive synced information about the music and even view a live orchestral score. The concert audience are invited to bring their mobiles and tablet with them to access this information whilst they enjoy the performance. The performances will all be streamed live online and will be available for 30 days after broadcast. We know we have a discerning audience who care not only about what we play but also about the way they hear it, how they experience it and how easy it is to hear it.

What we really celebrate this autumn, is not just about 70 years of the third programme or even how it became Radio 3, but what we celebrate is a pioneering spirit. A place and suite of experiences, funded by the licence fee, that allows everyone opportunities to understand humanity through pioneering music and culture - even nineteenth century music was shocking once.

As long as we continue to push frontiers both in commissioning and sound development, then we’ll be doing our job right for the next 70 years and continuing to enhance lives up and down the country.

Alan Davey is Controller of Βι¶ΉΤΌΕΔ Radio 3.

• .
• .

Today we’re launching another onto , giving audiences a unique Queen’s eye view of ‘Trooping the Colour’. Teams from around the Βι¶ΉΤΌΕΔ have been working on a small number of ‘true VR’ experiments as well, which we’re also unveiling today. Andy Conroy, who heads up Βι¶ΉΤΌΕΔ Research and Development, has blogged about some of these more advanced projects, which you can find on the . This is all part of the Βι¶ΉΤΌΕΔ’s early experimentation to help us better understand emerging technology and new mediums, explore the potential for future audiences, and see what kind of role the Βι¶ΉΤΌΕΔ should, or shouldn’t, be playing.

In terms of the technology, the main difference is that VR is more interactive and immersive than 360 degree video, which is very much as it sounds - a video played out in 360 degrees around the viewer.  However, both have the potential to give viewers a sense of presence. This is interesting to us as programme makers as it can help make people feel like they’re at the heart of the action or the story, which could help future audiences better understand important current affairs, news, science and history topics or give them a new perspective.

Experimenting with 360 video also has an advantage from a production point of view. The cameras or camera rigs are widely available and give programme makers the chance to see how it might complement our storytelling elsewhere, or get even more out of the projects they’re already working on. And we have hugely talented editorial teams inside Βι¶ΉΤΌΕΔ Studios wanting to explore this potential too - it's the people behind our BAFTA award winning TV series like Britain's Forgotten Slave Owners, VE Day, and Stargazing Live that are piloting these experiments on Taster.

But they are not doing this on their own. VR pioneers like Oscar Raby, digital agencies like Rewind, Aardman, VRTOV and Crossover Labs, Βι¶ΉΤΌΕΔ R&D specialists like Zillah Watson and Simon Lumb, along with teams in Βι¶ΉΤΌΕΔ Learning and innovation teams across Βι¶ΉΤΌΕΔ TV are forming a "coalition of the willing and able" to better understand the opportunities and challenges in this emerging field.

For example, as well as watching Alexander Armstrong's Βι¶ΉΤΌΕΔ One documentary Rome's Invisible City, Alexander can now take you on a . Βι¶ΉΤΌΕΔ Learning can complement a TV and Radio season around the Easter Rising by transporting you to O’Connell Street in 1916, immersing you in one of the Βι¶ΉΤΌΕΔ’s first VR documentaries . And in addition to watching Tim Peake on Βι¶ΉΤΌΕΔ Two's Stargazing Live, we can now put you in his spaceboots and send you 240 miles into the void above Earth in Βι¶ΉΤΌΕΔ – a VR Spacewalk - inspired by the real VR training Tim performed.

Audiences have already responded to other immersive pilots on Taster. In fact there have been over eight million views for projects as varied as the Natural History Unit's , Click's tour of the  and , all in 360 video.  Almost 2,000 pieces of feedback from people on Βι¶ΉΤΌΕΔ Taster have been completed with audiences responding favourably with an average score of 3.5 out of 5 stars, and comments stating that they felt immersed, part of the story - some even said it was "more real than TV".

Where these experiments with immersive technologies will take us is uncertain. They offer a glimpse of the promise of VR and what immersive experiences could provide to audiences, but they also bring new challenges we need to better understand. This is very early days, but these experiments, and the audience and industry feedback we can get from Βι¶ΉΤΌΕΔ Taster, are helping us do just that - helping inform any strategy the Βι¶ΉΤΌΕΔ may need in future.

Jessica Dromgoole is Editor of Βι¶ΉΤΌΕΔ Front, Radio 4's historical drama which spans the Great War. Here she explains how the programme is being used to launch the , an experimental online project on Βι¶ΉΤΌΕΔ Taster.

is a daily Radio 4 drama with a difference. It is a pithy and – I hope – delightful moment in a listener’s day. Each episode is set exactly one hundred years prior to broadcast, telling one fictional character’s story from that day with some genuine history thrown in too. Five characters’ stories are shared from Monday to Friday each week, with the collective played out each Friday evening as an .

Following the programme’s , Paul Donovan of the Sunday Times wrote “Βι¶ΉΤΌΕΔ Front is a poignant, quietly compulsive portrait of ordinary life a century ago, at once completely different and wholly recognisable”. We’re very proud of that, and what we’ve achieved with this programme so far, and hope to continue entertaining and educating our audience as the seasons pass.

Currently in Season 4, the project is planned to last 15 Seasons, a total of 600 episodes spanning the entire war, beginning on 4 August 1914 and ending on 9 November in 1918. We hope the sum of this work will be a partial account of ordinary people’s fortunes during the Great War.

Whilst the ambition is for the drama to feel organic; there are also strict timings to adhere to.

Each season is set in one place, and explores a different theme of the Great War. So far we have had the outbreak of war, the devastating slump in recruitment and the swathing industrial changes which followed. The current season touches upon profiteering and we’re currently working on Season 5, which charts the rise in spiritualism, and planning Season 6, looking at casualties and nursing.

I’ll forgive you if you haven’t quite kept up, but you can begin to get a sense of how big this project is.

Now consider staying on top of the family trees of EastEnders, or the last time there was a farming scandal in The Archers… Step forward .

Tracking information for programme makers has been an ongoing mission for the Βι¶ΉΤΌΕΔ Research and Development team. We were more than happy to be used as a guinea pig when they invited us to participate in an experimental online project which could not only go on to aid programme makers across the Βι¶ΉΤΌΕΔ, but also create something beautiful for Βι¶ΉΤΌΕΔ Front listeners to delve in to.

Via the Βι¶ΉΤΌΕΔ Front Story Explorer our regular listeners, and we hope new listeners, can navigate their own way through the fortunes of characters, locations, storylines, and national, international and local history in multiple dimensions.

Whether you’re most interested in one character’s progress through the war, a location that you may know, or a particular narrative, now you can trace their stories like threads pulled from the fabric of the whole.

With illustrations by Ivan Allen, and text by journalist Nick Curtis, it’s easy to click through the progress of the Βι¶ΉΤΌΕΔ Front narrative and listen only to the scenes that appeal. Links to small Βι¶ΉΤΌΕΔ features and blogs mean that the relevant historical research is also there at your fingertips.

The Βι¶ΉΤΌΕΔ Front Story Explorer is an example of the work Βι¶ΉΤΌΕΔ R&D are doing to explore new experiences that become possible as programmes are increasingly delivered over the internet.

For us it’s an incredible resource for our writers. It’s more than just an easy reference to the story that’s gone before, with every detail easy to locate. It’s the mulch from which new stories can grow, and an inspiration to keep Βι¶ΉΤΌΕΔ Front pithy and delightful. We hope our listeners enjoy using it as much as we do.

Jessica Dromgoole is Editor of Βι¶ΉΤΌΕΔ Front

• Βι¶ΉΤΌΕΔ Front Story Explorer can be found on .
• Βι¶ΉΤΌΕΔ Front is broadcast on Βι¶ΉΤΌΕΔ Radio 4, 12.04pm Monday to Friday, and episodes are available to listen again on .