Research Interests

My work primarily lies in the field of AI planning or automated planning, with a focus on developing innovative methodologies and systems. I often explore areas that are overlooked or underexplored in traditional planning-oriented venues such as ICAPS, aiming to bridge gaps between theoretical advancements and use-inspired research, as envisioned in Stokes’ Pasteur’s Quadrant. Below is an overview of the primary themes shaping my current research.

AI Planning

• Real-World HTN Planning. I am largely interested in Hierarchical Task Network (HTN) planning, focusing on enhancing it to address real-world challenges, such as complex constraints and risk-aware decision-making. My work examines what makes domains realistic and how such aspects can be integrated into HTN planning.
• Design-Driven Planning. I am also interested in methods, processes, and tools for designing and composing AI planning systems to align with specific requirements. The overarching question driving this research is whether tailored combinations of granular planning functionalities can produce planning systems that outperform off-the-shelf AI planners in solving complex planning problems.
• Sustainable AI Planning. I am actively working on understanding the energy consumption of AI planners, seeking to optimise their energy efficiency. While much of the field focuses on optimising planning time, my work considers environmental sustainability, recognising the growing importance of sustainable computing.
• Knowledge Engineering for AI Planning. I am also interested in exploring the use of Internet of Things (IoT) data and Large Language Models (LLMs) to automate the generation and refinement of domain models expressed in planning languages such as HPDL and PDDL.
• AI Planning Engineering. I am actively investigating the engineering of AI planning systems, including the specific software development lifecycle, provenance (data, workflow, metadata, system), and conceptual frameworks for domain modelling and system design. My work also examines experiences of AI planning practitioners.
• Applications. A significant aspect of my work involves applying AI planning, especially HTN planning, to real-world domains. The applications demonstrate the flexibility and transformative potential of AI planning in practical settings. Current areas of interest include: energy smart buildings, autonomous driving, distributed applications, robotics, and activity recognition.

Beyond AI Planning

Beyond AI planning, my research interests extend to automated service composition, where I focus on dynamically finding, selecting, and composing services to create adaptive and intelligent systems. I am also interested in learning algorithms from data, employing machine learning to derive domain-specific, efficient algorithms that enhance system capabilities.

Let’s Collaborate

If you find my research topics compelling and are interested in exploring potential collaborations, I would love to connect! I am particularly enthusiastic about interdisciplinary projects involving AI planning and its applications (excluding the military domain). Please feel free to reach out.