Edge computing has become an important way to process data and deliver intelligent services close to where data is generated, rather than sending everything to the cloud. Many edge systems already use machine learning to make predictions, detect anomalies, or monitor how a system is behaving. But these models usually stop at detection. They can flag a problem such as a resource bottleneck, a drop in performance, or unusual behaviour, but deciding what to actually do about it is still left to humans or to fixed, predefined rules. As edge environments grow larger and more dynamic, that approach becomes hard to scale and maintain.
This raises a natural question: can an edge system decide for itself when to move workloads, reallocate resources, or recover from a failure? Recent advances in Agentic AI offer a way to get there. Instead of passively predicting, agentic systems can reason about their environment, plan a suitable response, and adapt their behaviour over time. This project proposes a self-adaptive edge computing framework, built on Agentic AI, that moves beyond monitoring and prediction towards autonomous decision-making. The research will investigate how intelligent agents can continuously assess system conditions, coordinate their responses, and manage resources to make distributed edge systems more efficient, resilient, and adaptable.
Aim
The aim of this project is to develop an Agentic AI framework that lets edge computing systems move beyond passive monitoring towards making and acting on their own decisions. The project will investigate how intelligent agents can understand current system conditions, make sensible decisions, and coordinate actions across distributed edge resources. Through this work, the research aims to establish new ways of building self-adaptive edge systems that can operate more independently, efficiently, and reliably when conditions are constantly changing.
Objectives
This project will first examine the limitations of current AI-enabled edge management, which can detect and predict but rarely act, and identify where greater autonomy is possible. It will then design an Agentic AI architecture that can monitor system conditions, reason about operational goals, and choose appropriate actions as circumstances change. The research will explore mechanisms for adaptive resource management, workload optimisation, and fault recovery, and study how multiple agents can cooperate to support system-wide goals rather than acting in isolation. Finally, the framework will be evaluated using realistic edge computing scenarios to assess its effectiveness, scalability, and ability to improve operational efficiency compared with conventional rule-based or prediction-only approaches.
Significance
Edge systems are becoming more intelligent, but they still mostly rely on passive prediction and monitoring. This project explores a shift towards self-adaptive edge infrastructure that can make decisions and adapt on its own. By bringing Agentic AI together with edge computing, the research has the potential to improve how resources are used, reduce the operational burden on human operators, and make systems more resilient when conditions change unexpectedly. The project will contribute new knowledge at the intersection of Agentic AI, distributed systems, and edge computing, while addressing an emerging challenge for next-generation digital infrastructure. Its outcomes are expected to support the development of more adaptive, self-managing computing environments that can operate effectively in complex and fast-changing conditions.
Ideal Candidate
We are looking for a self-motivated PhD candidate with strong analytical, problem-solving, and research skills. Candidates with a background in Computer Science, Software Engineering, Artificial Intelligence, Data Science, or a related field are encouraged to apply. Experience with machine learning, distributed systems, cloud or edge computing, Python programming, and software development is desirable. Familiarity with intelligent agents, large language models, reinforcement learning, or autonomous systems would be an advantage. Additionally, the applicants should meet the eligibility criteria for entry into a PhD program at Curtin University.
This project is open to International and Domestic applicants.
Internship
This PhD project may offer the opportunity for an industry internship during the candidature, providing real-world research experience and exposure to applied AI in an operational setting. There is potential for the internship to be undertaken with an industry partner such as Thales or L3Harris, subject to availability and mutual interest. An internship of this kind would allow the candidate to apply their research on self-adaptive, agentic edge systems to practical, real-world deployment challenges, while building valuable industry connections. The specific scope, host organisation, and timing would be confirmed during the candidature.
Scholarship
If you are identified as the preferred candidate for this project, you may be considered for an RTP scholarship.
Enquires and How to Apply
For enquires about this opportunity contact Associate Professor Sajib Mistry at Sajib.Mistry@curtin.edu.au
To formally apply submit an Expression of Interest to Associate Professor Sajib Mistry during the Central Scholarship round (July 1st – July 31st 2026)