enableIT: Technology based Inclusion via Human-Centred System Analysis and Assistance in Industrial Environments
Digital assistance systems for the support of manual assembly and packaging activities are increasingly used in the industrial environment. Especially in inclusive contexts, these systems have proven to be beneficial in supporting people with cognitive impairments in carrying out variant- rich processes. The systems provide processinformationandguidelinesin an interactive way and thus provide cognitive relief. However, current systems usually only offer a static form of assistance. This results in a limited user experience, as the information provided is not adapted totheindividualabilitiesandneedsof the workers. Within the scope of the project, an assistance system will be developed that automatically adapts to the abilities and needs and the learning progress of the workers.
Previous studies have shown that the content and the scope of the provided assistance should be adaptable to the individual abilities and needs of the user. However, there is no evidence so far on how changes in the offered assistance have to be implemented in order to achieve an optimal user experience. To facilitate an appropriate classification of the learning progress, methods from the field of Artificial Intelligence (AI) are to be used. However, models must be developed to correctly represent the particular circumstances. Furthermore, concepts for the adaptation of the provided assistance and its communication to the user have to be developed.
In the course of the project two projection-based assembly assistance systems have been developed, which will be integrated into the production environment of the Integ GmbH. The assistance software has been extended to include functions for the gradual adaptation of the assistance. In the course of an initial study, the effects of dynamically modified assembly assistance on assembly times and error rates as well as acceptance are being analysed. By using an eye-tracking system to record eye movements, the actual use of the provided assistance information will also be investigated. The data obtained from the study will further be used to develop AI- based models that can be used to automatically determine the learning progress and the adaptation of the assistance.
² Last authors
² Last authors