All the bbc.co.uk /programmes and /music RDF metadata is being crawled into a Talis data store, which is nice.
You can do things like this with it:
to select all episodes of each series of a specific programme.
PREFIX po: <http://purl.org/ontology/po/>
PREFIX dc: <http://purl.org/dc/elements/1.1/>
SELECT ?episode ?title ?synopsis ?series_title ?programme_title
<http://www.bbc.co.uk/programmes/b00k3wvk#programme> dc:title ?programme_title;
?series dc:title ?series_title;
?episode dc:title ?title.
?episode po:medium_synopsis ?synopsis.
You can use that query in a Pipe too, to create a badge or RSS feed. Yahoo Pipes doesn't have a "SPARQL builder" module yet, but you can use the String Builder to create a SPARQL query.
Then what? In theory, you should be able to pass the URI of each item to URIplay and receive information on where to retrieve the actual programme data, then play it.