at footage from _ performances. so we are looking at footage from different - at footage from different sports, subtracting and extracting movement out of it frame by frame, we are then doing pose estimation, which essentially draws out different elements of that movement, and then applying that to an algorithm that then drives our artistic representation of that in the robot's movements. the art installation _ in the robot's movements. the art installation was _ in the robot's movements. the art installation was shipped from the uk right into the heart of tokyo to create over 150 unique illustrations to run alongside the games until september. we alongside the games until september.— alongside the games until september. we are working mainly from _ september. we are working mainly from audio - september. we are working mainly from audio clips, - september. we are working| mainly from audio clips, but september. we are working - mainly from audio clips, but we have also worked with some numerical data, results data, be that speeds or times or heights or distances. i think there is no _ heights or distances. i think there is no doubt _ heights or distances. i think there is no doubt that - there is no doubt that technology opens up of potential in the but one of the bil potential in the but one of the big barriers is that you can access_ big barriers is that you can access the technology in the first — access the technology in the first place, be it hardware or software _ first place, be it hardware or software. a lot of the stuff that — software. a lot of the stuff that is _ software. a lot of the stuff that is needed for that is actually being generated in the commercial sector in film, in