In engineering, robustness is defined as the “degree to which a system or component can function correctly in the presence of invalid inputs or stressful environmental conditions”. In this project, we will focus on the robustness of Generative AI (GenAI) – the new and powerful class of AI models, which have been revolutionizing the way complex systems are built, based on their ability to recognize patterns, make predictions, and even create new content. The unprecedented complexity and broad scope of GenAI systems require revisiting the engineering-focused notion of robustness. In particular, GenAI systems engage directly with a wide spectrum of users in a broad range of environments and produce outputs that can significantly affect these users and, more broadly, the society. This calls for a robustness framework that goes beyond technical robustness and encompasses resilience to user diversity and considerations of ethical fairness. Moreover, the need to work in an open and dynamic environment requires the capacity to manage diverse and unforeseen scenarios. By adopting a broader view of robustness, novel techniques can be developed to ensure that GenAI systems are not only technically sound, but also socially responsible, reliable, inclusive, and capable of maintaining high-quality performance across a wide array of environments and interactions.
PPM Robotics AS will especially be involved into R&D on utilization of AI to create a natural interaction and communication between elderly with dementia, and service-robots operating in nursing home and home care.
The outcome is intended to be using in PPM's future technology, as a part of our strategy to become a leading provider of social service robots for nursing homes and home care of elderly.