Two-day Workshop: Verification and Validation of Autonomous Systems: Ethical, Social and Trustworthy behaviour

Workshop venue: Cumberland Lodge, Windsor Great Park

Background

Autonomous systems are on the increase across society, from driverless cars and unmanned air vehicles, to web-bots and exploration robots. The core aspect of an “autonomous system” is that it makes decisions for itself, about what to do, when to do it, where to do it, etc. But since these decisions are now taken away from humans, how do we know the autonomous system will make the choices we would have done?

Increasingly, autonomous robots are designed to interact with, take care of safety for, and even be trusted companions of, people. This class of autonomous system requires important work on

– their social aspects, i.e. how then interact with people,

– their trustworthiness, i.e. what will convince us to trust them, and

– their ethical aspects, i.e. what moral decisions might they have to make.

There is much work on these aspects across the world but a key aspect that must also be covered is how can we really ensure that our systems behave as expected. It is here that various forms of verification and validation, from formal proof, through testing, to user validation.

Workshop

This workshop, held on 23-24 November 2017, brought together researchers who are developing ethical, social and trustworthy behaviour in autonomous robotics, with researchers across the range of verification and validation techniques, in an informal setting to discuss these issues, to share understanding, and to stimulate collaboration. The workshop was funded by EPSRC, through both the Network on the Verification & Validation of Autonomous Systems, and the Verifiable Autonomy research project. It was chaired by:-

Structure

The workshop comprised of a range of invited talks from both the UK and across the world, together with discussion sessions and proposals.

Workshop programme

Presentations:

 

Advertisements