 We're here at the Linaro Connect, a Budapest, and who are you? Hey, I'm Syed Mohammad Khasim. I work for Linaro in Android TV, a project. So we have this new project called AOSP TV, which is here. Android AOSP TV. It sounds very interesting. Yeah, it's an AOSP TV lead project from Linaro. So the goal here is to get all the AOSP components for TV from AOSP. Get all the Android TV components from AOSP. And then we build all the other differentiators into that, like integrating our Opti security framework. And then we integrate all the open applications which are available in the open source, like AXO player and Android sample inputs, live TV apps and these things. And then we provide one end-to-end solution for our members, which is called MLCR from Linaro Home Group, which is meant for smart TVs and setup box to just enable the premium services around the Android TV segment. So here we are showing a demonstration, which is like running Android TV on Haiky. So we will be replacing Haiky with one of the Android TV platforms from 96 boards in future, but in order to enable some cool demos, so we started off with Haiky today. So right now it's playing a video. How does the UI look like? The UI looks good. We have not enabled any media accelerators yet. It is all software decoding, which is happening on Haiky. But in future, when we get to the TV platform, we will be enabling media accelerators along with the Android TV components. So I have with me Vincent Shaw. He's from MSTAR and I have Shalu. He's from Shoshonex. So they are working with me in my team doing all the Android TV stuff. So they are mainly the security guys who are helping me with building all the trusted applications for Widevine and Plarity and also integrating the Opti framework with Android TV. So we will be having an end-to-end solution there. So it's basically DRM stuff, right? Yes. So how is it working? Basically we are integrating Opti with YY's DRM technologies and also we can have some media playback for secure things. So there's a logo here showing Netflix and Amazon and some other requirements for specific DRM support, right? Yeah. And this is what you're trying to make work with the open source AUSP TV. Yes. And how does that work? How is it possible to do a DRM in open source? I mean, I'm joking. Yeah, it is possible. Android comes with all DRM stuff. So what we enable is the secure side of things with the Linaros Opti framework, which will actually give the... Like we write the trusted application on the security side and we use all the existing DRM framework and as I said, the media accelerators are not in there. So when we do complete media accelerator integration it will become like an end-to-end secure media playback. So that is the goal of this project to give end-to-end solution. So it's going to support hardware acceleration with DRM somehow. Connect from one to the other and you're able to do this. It's working. We are still working on this part because we have the secure processor and we can switch the secure wall to perform the question and we can keep the career content data in the secure memory and the NINAS kernel cannot access the secure memory. So we can isolate both the NINAS kernel and the secure environment. And so this is quite important. I mean, when I go to China, there's a lot of Android TV boxes but what are they running? Some kind of messy Android, no? Like just some Android from the smartphone they put on a TV box. But this is not as good as... Yeah, I don't know what they're running so I cannot comment on that. But definitely if you want to have GMS, Google applications running on your box then you need to get official certifications from Google so there is a complete process that you need to follow. What we give from Linaro is an intuitive solution with the CTS compatibility so we test the software for CDD compliance so with this software and adding any custom UI and adding the differentiators our members can go to Google with an end solution for getting easy certifications so that will help them to quickly get any kind of certification and GMS. So you work in the open source Linux world, the open source Android and you've been doing Android in Linaro for a while, right? Yeah. So yeah, personally I was part of Android core working group in Linaro and then I was doing all Arm V8 stuff for Linaro so once Arm V8 stuff was done 64-bit Android and other things were done so then I moved to... I was working in different projects now I'm doing Android TV for Linaro. Your work is basically in the pocket of everybody in the world, right? I think so. Because your software is going back into Android everybody's using it. Is Google making Android TV open source or you just say we work in open source we do whatever we want and you start doing even before Google puts it open source or what is it? No, Android TV Google already has everything in open source like AOSP project is completely open source. For the whole Android TV? Yeah, the TV input framework everything is already open source but what we need to do is the hardware abstraction layers for linear devices like antennas or like your HDMI in or video in these kind of things so they need a hardware abstraction layer enabled so those things we add as some live examples on our 96 board TV platform because they have HDMI in and they have a few video in there and like antenna input devices so we enable those input devices along with the IPTV layer as well so it is media over IP as well as media over linear devices so we enable both of these. So how is Android TV different from Android what is just the part it's kind of like one type of Android? So mainly there is very limited difference between Android for mobile and Android for TV but the main differentiation factor here is the TV input framework which Google has provided which is very good and very which gives a clear layer for integrating any TV input devices so that is very good and very neatly done by the Android team at Google but as I said like the linear input devices have to be integrated by the vendor and the hardware abstraction layers have to be implemented and the TV input service should be should be integrated with these hauls and then the live TV channels and live TV app which again Google provides so those things can show up these channels and user can select one of these channels to watch the TV so that way it is very clear for Android TV developers on what path or what framework to exactly change and they don't have to work across all the Android components of Android as such So for example just with M-Star chips there is many smart TV, many TV box do they run many different type of Android out there is it messy and linear is trying to make it better and more smooth right? as the work is happening here right? I think so for now M-Star is using the proprietary Android trunk but we are working with Android and we hope consider using something different like something more open or something what's called more optimized everything better right? so the Android TV experiences will not be fragmented but actually and there will be a big part that means your work is not only in everybody's pocket but it's going to be in everybody's living room yeah so for our members the main value add that we are adding like for M-Star or High Silicon or for the Shoshone X or any so our work is for SOC vendors our work is for content providers our work is for community application developers and the OEMs and ODMs so the work that we are doing as part of MLCR for Android TV it's actually to cater to all these various different end customers or end users so what they get is like an end to end solution of Android and they don't have to really work on the baseline components so baseline components are already developed by us which includes the secure data path secure media playback the complete Android TV input framework with all the linear devices and IPTV devices so all these things are part of our starter kit kind of thing with the 96 board reference platform for TV which is based on the 96 board spec for TV so all these things come as a bundle thing to the end consumer and then they can actually work only on the key differentiating factors like their accelerators or codec accelerators and doing any kind of custom UI or regional specific content that if they want to populate so all these things can be done by the SOC vendors and content providers and they are not they don't have to get distracted with all the baseline software which we have already delivered it to them so they will just be able to ship very good hardware acceleration of all the video and also DRM so they will potentially be more 4K Netflix 4K and Amazon 4K because they will be ready to be approved that it is secure and smooth and fast and everything. Our initial goal is to get wide wine working and then we already had worked on play ready integration with Android so those two things are quick we should be able to provide quickly and also clear key we can make clear key validation also so all these things will is something which we are working towards and everybody who is making an SOC for TV box should work with you right everybody should join you want to have Amlogic Rockchip some other At this point we are really not looking at very low cost segment because that is very competitive and the challenge there is like getting time to market faster and the software experience there will be limited because the SOC that they will be using is not so premium but wherever it is high end setup box or advanced setup box with the latest features of Android and wherever it is with the extended features and functionality there Android TV would really look amazing compared to any other OS that is available for TVs today so how do you feel about working with all these people and the narrow it is fantastic and from different companies everybody join together work together we have common goals and we can have achieve some tasks and it is fun fun work cool thank you