The journal Entropy is soliciting papers for a Special Issue on emerging applications and theory of information theory, as applied broadly in the field of robotics and control. Both application-driven research, cultivating and promoting non-conventional uses of information theory in robotics and control, as well as theory-oriented research papers in these areas are solicited.
Recent advances in autonomous systems, have made it clear that a key aspect to the successful development of these systems is the harmonious integration of a diverse set of disciplines including perception, cognition, control, decision-making, and planning, among others. Motion planning and control techniques, in particular, have been bolstered in recent years partly because of the advancement of capable computational platforms, the availability of low-cost sensors, combined with the prevalence of statistical (machine) learning techniques and methodologies, that have allowed operation in poorly characterized or previously unknown environments. Both in motion planning and control, a fundamental issue is uncertainty characterization and uncertainty mitigation using feedback. There is a growing realization in the community that information theory can play a larger role in this context, as it can provide the correct framework, along with the right set of tools, to answer important questions such as what is relevant in the problem and what is not, what is the best way to transmit information between the controller and the sensor, what signals to communicate between various agents in a multi-agent network to manage bandwidth and/or mitigate external attacks, etc. Information theory can provide the missing link to close perception-action-communication (PAC) loops in complex autonomous systems. There is a growing body of the literature, where information-theoretic concepts play a role in several contexts, including state representations, strategic perception, communication and coordination in multi-agent systems, and the analysis of machine learning algorithms.
Topics relevant to this Special Issue include (but are not limited to):
• Intelligent perception
• Information theory in reinforcement learning
• Multi-agent and networked control systems
• Information-theoretic state representations
• Statistical mechanics in control and decision-making
• Joint communication, sensing, and control
• Resource-constrained control, planning, and perception
• Entropy and feedback systems
The Special Issue will continue to accept submissions until June 15, 2021. All papers will be published immediately after acceptance on an ongoing basis. For more information please consult
https://www.mdpi.com/journal/entropy/special_issues/motion_control
Takashi Tanaka and Panagiotis Tsiotras (Guest Editors)
