Professor Inna Skarga-Bandurova

Mechanisms of Autonomous Decision-Making in Self-Governing Robotic Systems: Provisional Results and Lesson Learned

Autonomy represents self-governing behaviour, including independent decision-making, adaptive navigation, and real-time environmental perception. These are crucial components for systems dedicated to improving safety and security.

Self-aware autonomy requires precise information processing at three levels: perceiving scene elements, understanding the current situation, and projecting future status. Autonomous robots and unmanned vehicles depend on advanced algorithms and sensor systems for independent navigation in complex environments, making instant decisions, and adapting to dynamic scenarios. Nonetheless, universal solutions are still in their early stages and demand ongoing exploration and improvement.

This presentation delves into the significance of autonomy in constructing reliable and resilient robotic systems while exploring the core principles of autonomous decision-making. Our discussion encompasses ongoing research, including high-level decision-making policies grounded in complex activity recognition, illustrated with practical examples drawn from semi-autonomous robotic platforms. We also share insights gained from testing these decision-making systems, shedding light on their potential and constraints in selecting appropriate actions. Furthermore, we showcase the effective utilization of diverse data sources in aiding robot-assisted tasks.

Bio:

Professor Inna Skarga-Bandurova

iskarga-bandurova@brookes.ac.uk

E209, Wheatley Campus, Wheatley, Oxford, OX33 1HX

Affiliation:

• Oxford Brookes University, Oxford, UK
• Ternopil Ivan Puluj National Technical University, Ternopil, Ukraine
• G.E. Pukhov Institute for Modelling in Energy Engineering, Kyiv, Ukraine

Inna Skarga-Bandurova is a Senior Lecturer at Oxford Brookes University and a Visiting Professor in the Cybersecurity Department at Ternopil Ivan Puluj National Technical University, as well as the Department of Mathematical and Econometric Modelling at G.E. Pukhov Institute for Modelling in Energy Engineering. With extensive experience, she serves as a senior research scientist, tech lead, and project manager, co-leading teams focused on data analytics and responsible AI.

Her expertise spans foundational research, product development, and implementation, particularly in uncertainty modelling and decision theory. Her research group specializes in developing mathematical techniques and toolkits for explainable AI and verifying neural network behaviour, offering dependable solutions across various domains, including assistive robotics, cybersecurity, environmental studies, and medical research.