This talk will provide an overview of some projects related to accessibility which have inspired new approaches to audio description research. First of all, I will refer to the VIW project, in which a multilingual corpus of audio descriptions was created, allowing a cross-linguistic and intralinguistic analysis. One of the possible analyses concerns audio description voicing, an aspect that has been investigated as part of the RAD project. Furthermore, using the results of the EASIT project, it will be discussed if existing audio descriptions could (or should) be made easier to understand, as well as how collaborative processes related to artistic co-creation could have a place in audio description practice (TRACTION project), and how copyright aspects could be managed using blockchain technology (Mediaverse project). The ultimate aim is to approach audio description research from a myriad of perspectives that allow to engage in a fruitful dialogue on where audio description research is heading to.