Newcastle University is offering one fully funded, 4-year PhD studentship in Computer Science focused on Autonomic Computing and Agentic AI for the Internet of Things (IoT) and Edge Systems. This is an EPSRC-sponsored opportunity with 100% tuition fees covered and a minimum tax-free annual living allowance of £20,780 (UKRI rate for 2025/26), plus additional project costs.
If you’re interested in building self-managing, resilient, intelligent edge devices—where compute, memory, connectivity, privacy, and energy are all constrained—this project is designed around exactly those real-world challenges.
Author: Dr Niaz Chowdhury (LinkedIn)
Designation: Lecturer (Computer Science)
Affiliation: Ulster University, Birmingham, UK
Key details at a glance
- University: Newcastle University
- Programme: PhD (Computer Science)
- Project title: Autonomic Computing and Agentic AI for the Internet of Things and Edge Systems
- Location: Newcastle upon Tyne, UK
- Funding eligibility: UK students, EU students, International students
- Funding package:
- 100% fees covered
- £20,780 minimum tax-free stipend per year (2025/26 UKRI rate)
- Additional project costs provided
- Start date: 1 October 2026
- Duration: 4 years
- Number of awards: 1
- Closing date: 15 February 2026
- Reference: DLA2632
- Supervisors: Dr Tomasz Szydlo, Dr Devki Nandan Jha, Dr Rishad Shafik
- Contact: Dr Tomasz Szydlo
Why this PhD matters now
IoT and edge computing are shifting from “collect data → send to cloud → analyse centrally” to something much more ambitious:
- Distributed intelligence happening on-device
- Decision-making at the edge (not always in the cloud)
- Adaptive behaviour under changing environments and unreliable connectivity
But edge systems face tough constraints: limited compute and memory, intermittent networks, strict privacy requirements, and tight energy budgets. Under these conditions, the project highlights a key idea:
Edge devices must manage themselves.
That is where autonomic computing comes in—systems that can self-configure, self-optimise, self-heal, and adapt without constant external control.
The research focus: Agentic AI meets Autonomic Computing (at the extreme edge)
This PhD investigates the convergence of:
- Agentic AI (agents that can perceive, reason, plan, act, and learn), and
- Autonomic computing (self-managing behaviour: configuration, optimisation, fault recovery, adaptation)
The goal is to design and deploy genuinely self-managed and resilient IoT/edge systems, suitable for real application domains such as:
- Smart buildings and smart cities
- Healthcare devices
- Industrial IoT
- Connected and autonomous vehicles
A core challenge is that many classical AI and distributed systems approaches do not translate directly to the edge, because edge devices are severely resource-limited.
What problems you might work on
Based on the project scope, your work is likely to explore questions such as:
1) How can edge agents adapt continuously in dynamic environments?
Agents need to update their behaviour and knowledge while operating under constraints (compute, memory, energy) and changing real-world conditions.
2) How do we keep edge decision-making explainable?
As systems become more autonomous, explaining why an agent took an action becomes important—especially in safety- and trust-critical contexts.
3) What lightweight learning methods work best at the extreme edge?
The project motivates lightweight, logic-based machine learning approaches to support adaptation and explainability in constrained environments.
4) How can multiple edge agents collaborate without relying on the cloud?
A major theme is decentralised collaboration: agents sharing observations, learned models, or inferred knowledge to coordinate, detect anomalies, and optimise system-wide behaviour—while staying:
- Lightweight
- Scalable
- Privacy-aware
- Resilient to intermittent connectivity
Why now: enabling technologies that make this feasible
The advert notes recent advances that have made sophisticated on-device reasoning more realistic, including:
- Small language models
- TinyML frameworks
- Sparse neural networks
- Microcontroller-grade accelerators
These developments open the door to deploying more capable intelligence directly on constrained devices—alongside their primary applications.
Supervision and research environment
You will be trained and supported by experts in the Intelligent Systems Research group, with experience in:
- Deploying machine learning models on edge devices and microcontrollers
- Building simulation environments for IoT
The project also includes collaboration with:
- Literal Labs (a Newcastle University spin-off working on logic-based ML models)
- National Edge AI Hub
- An extensive international partner network
This combination is particularly valuable if you want both strong academic supervision and exposure to applied, real-world edge AI development.
Who should consider applying?
This PhD is a strong fit if you’re excited by topics such as:
- IoT and edge computing architectures
- Distributed and decentralised systems
- Agent-based AI and autonomous decision-making
- Fault tolerance, resilience, and self-healing systems
- TinyML / on-device learning / resource-aware ML
- Explainable AI (especially under constraints)
- Privacy-preserving, intermittent-connectivity-aware collaboration
You don’t need to match every area—what matters most is a strong alignment with the central challenge: building self-managed intelligence at the edge.
How to apply (and what to prepare)
The advert states that eligibility criteria and application instructions are provided on Newcastle University’s website. In practice, for a competitive fully funded PhD like this, it’s wise to prepare early and ensure your documents clearly reflect the project’s themes.
Application checklist (recommended)
- CV highlighting relevant technical experience (projects, research, publications if any)
- Academic transcripts
- Personal statement / research motivation statement aligned to the project scope
- Evidence of relevant skills (e.g., embedded systems, ML, distributed systems, simulation, IoT toolchains)
- References (as required by the university process)
A strong “fit statement” should show:
- You understand why edge constraints change everything (compute, memory, energy, privacy, connectivity)
- You can connect your experience to autonomy + resilience + collaboration
- You can articulate interest in lightweight and explainable approaches (not “big cloud-first AI”)
Deadlines and timeline
- Advert placed: 23 January 2026
- Application deadline: 15 February 2026
- PhD start date: 1 October 2026
- Duration: 4 years
With a deadline in mid-February, aim to have your key documents ready well before the closing date.
Quick FAQ
Is this fully funded?
Yes. It covers 100% fees plus a minimum £20,780 annual tax-free stipend (2025/26 UKRI rate), and includes additional project costs.
Can international students apply?
Yes—funding is listed for UK, EU, and International students.
How many awards are available?
One studentship (highly competitive).
Who can I contact with questions?
Dr Tomasz Szydlo is listed as the contact.
Final note
If you want to work at the frontier of self-managing IoT/edge intelligence, where systems must operate reliably without constant cloud dependence, this Newcastle University PhD is a rare and timely opportunity—especially with full EPSRC funding and a supervision environment strongly grounded in real edge deployment.
Reference to quote in your application or email: DLA2632.
