Scrutable Plan Enactment via Argumentation and Natural Language Generation

Autonomous systems suffer from opacity due to the potentially large number of sophisticated interactions among many parties and how these influence the outcomes of the systems. It is very difficult for humans to scrutinise, understand and, ultimately, work with such systems. To address this shortcoming, we developed a demonstrator which uses formal argumentation techniques, coupled with natural language generation, to explain the rationale of a hybrid software-human many-party joint plan during its enactment.