Integrated Individualized Modeling towards Cognitive Control of Human-Machine Systems

Northeastern University, Boston MA

Investigators

Abstract

The research objective of this project is to integrate human factors and control for modeling of collaborative human-machine systems. Specifically, in a driver-vehicle system, the driver's cognitive state and behavior may become significant factors to the overall system performance. This work is to study a driver assistance system (DAS) that integrates the driver's cognitive state and behavior, as well as the vehicle state and behavior, as inputs. Driver's cognitive state includes emotion, mental workload and attention. The approach has three main aspects: (1) investigate a method to model a driver's cognitive state and behavior while controlling a vehicle in a dynamic human-machine environment (HME); (2) incorporate the HME model into a DAS as a form of feedback control; (3) test and validate the above theories through the DAS simulation. The research makes significant contributions to the areas of human state and behavior modeling, as well as cognitive control of human-machine systems. Driver assistance systems have a strong relevance to transportation safety, and one large deliverable from this research includes a reduction in automobile accidents. This DAS is a specific test bed for a more general Human Assistance System (HAS). The test bed results can be used to model a more generic HAS, which can then be applied to human-machine systems in a variety of domains such as medicine, manufacturing, and space applications. This work will provide experience to students and teachers by providing educational activities to support k-12 education. Graduate and undergraduate students will benefit through research integration with curriculums and senior design projects. This work expects to attract minorities and underrepresented groups, especially women. The work will also reach out towards professional societies and international collaboration for a wide dissemination of the research outcome.