Using Learning Objective-based Course Modeling for Complete Exercise Generation: From Course Material to an Aggregated Knowledge Representation

In recent years, mobile robotics applications have drastically grown in complexity not only with respect to the tasks itself but also in the area of application the robots are deployed. This development not only comes with an increased complexity of the robots themselves but also with a higher risk of hazardous events caused by unforeseen situations or internal faults. Previous work has focused on challenges that arise in unstructured and uncontrollable environments such as public roads and sidewalks and on technologies that provide tolerance to faults in low level functionality. However, ways to cope with failures in the robot's high level control system have largely been overlooked. In this paper, we argue for fail-operationality in high level control systems as one vital characteristic of safety in mobile robots. We then present a novel control architecture that allows for state machine replication which ultimately ensures fail-operationality with respect to internal faults not only in low level functionality but also the control architecture itself. The architecture was implemented, tested and its fail-operationality validated experimentally. The concept presented in this paper provides an infrastructure that allows for the integration of a new range of safety technologies in autonomous mobile robots.

DDC Class

629.8: Control and Feedback Control Systems

005: Computer Programming, Programs, Data and Security

006.3: Artificial Intelligence

620: Engineering

Publication version

publishedVersion

Lizenz

http://rightsstatements.org/vocab/InC/1.0/

Name

TALE_2023-1.pdf

Type

Main Article

Size

300.87 KB

Format

Adobe PDF

Options

Using Learning Objective-based Course Modeling for Complete Exercise Generation: From Course Material to an Aggregated Knowledge Representation