From training to robot behavior: Towards custom scenarios for robotics in training programs for ASD

Successful results have been booked with using robotics in therapy interventions for autism spectrum disorders (ASD). However, to make the best use of robots, the behavior of the robot needs to be tailored to the learning objectives and personal characteristics of each unique individual with ASD. Currently training practices include adaptation of the training programs to the condition of each individual client, based on the particular learning goals or the mood of the client. To include robots in such training will imply that the trainers are enabled to control a robot through an intuitive interface. For this purpose we use a visual programming environment called TiViPE as an interface between robot and trainer, where scenarios for specific learning objectives can easily be put together as if they were graphical LEGO-like building blocks. This programming platform is linked to the NAO robot from Aldebaran Robotics. A process flow for converting trainers' scenarios was developed to make sure the gist of the original scenarios was kept intact. We give an example of how a scenario is processed, and implemented into the clinical setting, and how detailed parts of a scenario can be developed.