The Fostering Attachments Group (Golding, 2006) is a group intervention combining social learning theory and attachment theory to inform the parenting of looked after children who present a wide range of emotional and behavioural difficulties. Evaluations to date have consistently found high rates of satisfaction from participants and a greater understanding of the child's needs, but only limited impact on the child's behaviour. Identifying the mechanisms of change in these well-received groups, therefore, seems important. Ben Gurney-Smith, Charlotte Granger, Anna Randle and Jenny Fletcher report on a novel evaluation of this group using, for the first time, a mixed group of foster carers and adoptive parents, with measures repeated at pre-, post- and three- month follow-up, and consistent with both theoretical models. The group was rated highly for satisfaction and sustained improvements were found in the caregiver's understanding and the perceived difficulties of the child. Significant positive and sustained changes were found in specific behavioural difficulties named by the caregiver, the child's hyperactivity and in the caregiver's mind–mindedness when they perceived a break in the relationship with their child. The study supports the use of measures capturing change over time in both the child and caregiver, which are consistent with both theoretical approaches used within the group.

Autonomous driving promises to transform road transport. Multivehicle and multi-lane scenarios, however, present unique challenges due to constrained navigation and unpredictable vehicle interactions. Learning-based methods-such as deep reinforcement learning-are emerging as a promising approach to automatically design intelligent driving policies that can cope with these challenges. Yet, the process of safely learning multi-vehicle driving behaviours is hard: while collisions-and their near-avoidance-are essential to the learning process, directly executing immature policies on autonomous vehicles raises considerable safety concerns. In this article, we present a safe and efficient framework that enables the learning of driving policies for autonomous vehicles operating in a shared workspace, where the absence of collisions cannot be guaranteed. Key to our learning procedure is a sim2real approach that uses real-world online policy adaptation in a mixed-reality setup, where other vehicles and static obstacles exist in the virtual domain. This allows us to perform safe learning by simulating (and learning from) collisions between the learning agent(s) and other objects in virtual reality. Our results demonstrate that, after only a few runs in mixed-reality, collisions are significantly reduced.