The AVA+ project aims to study the application of virtual animated assistants for visitor support in public spaces, reducing staff workload. Virtual assistants can be deployed across a fleet of terminals, either stationary or mobile, and "jump" between them while maintaining conversation context.
In the prototypical AVA+ use case, a user arrives at a venue, e.g. a museum, and heads to a virtual assistant terminal at the entrance. After being asked about an exhibit, the assistant summons a mobile robot to guide the user to its location (a). When the user arrives, the assistant resumes the original conversation about the exhibit (b).
<iframe width="560" height="315" src="https://www.youtube.com/embed/8HCWuq3J-ag?si=BcaL_H7mveU4FNej" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>Robot guiding user to washroom
<iframe width="560" height="315" src="https://www.youtube.com/embed/VxArQkt0M4s" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>The conversational context of AVA+ virtual assistants is maintained in backend servers. Terminals select a specific context based on identification data (voice, visual appearance, etc) collected at the beginning of a session. Additionally, mobile platforms receive navigation commands, which are executed by a local autonomous navigation stack.
