# Research: UCD Linked Agents and Intelligence Research Group

<a id="multi-agent-systems"></a>

## [Agent-Oriented Programming & Multi-Agent Systems](research/multi-agent-systems.md)

Agent-oriented programming treats software as a collection of autonomous agents that reason about beliefs, goals and plans. We design and implement languages and runtime platforms that make these ideas practical for real software engineering, with a focus on the ASTRA language and its toolchain.

Multi-agent systems bring together many agents that coordinate, cooperate or compete. We investigate the abstractions, organisational models and software engineering practices needed to design, test and deploy such systems reliably.

**People:** [Rem Collier](people/rem-collier.md), [Katharine Beaumont](people/katharine-beaumont.md), [Hao Mu](people/hao-mu.md), [Mahjabin Siddique](people/mahjabin-siddique.md), [Md Asif Iqbal Fahim](people/md-asif-iqbal-fahim.md), [Eoin Hobson-O'Neill](people/eoin-hobson-oneill.md), [Vivek Nallur](people/vivek-nallur.md)

**Keywords:** ASTRA, BDI, AgentSpeak, Agent Platforms, Coordination, Organisations, Testing

<a id="hypermedia-mas"></a>

## [Hypermedia & Web-based Agents](research/hypermedia-mas.md)

The Web offers an open, uniform environment for agents to perceive and act. We explore how agents can navigate hypermedia, use semantic descriptions of resources, and interoperate with Web of Things devices.

**People:** [Rem Collier](people/rem-collier.md), [Katharine Beaumont](people/katharine-beaumont.md), [Eoin Hobson-O'Neill](people/eoin-hobson-oneill.md)

**Keywords:** Hypermedia, Linked Data, Web of Things

<a id="llm-agents"></a>

## [Agentic AI & Language Models](research/llm-agents.md)

Large language models have renewed interest in autonomous agents. We study how LLMs can be integrated with explicit reasoning, planning and multi-agent coordination to produce agentic systems that are more capable, predictable and explainable.

**People:** [Rem Collier](people/rem-collier.md), [Katharine Beaumont](people/katharine-beaumont.md), [Vivek Nallur](people/vivek-nallur.md)

**Keywords:** LLMs, Neuro-symbolic, Explainability

<a id="social-simulation-ethics"></a>

## [Social Simulation & Machine Ethics](research/social-simulation-ethics.md)

We build multi-agent models of how people and machines make decisions under uncertainty, from individual factors such as cognitive biases and emotions to loosely coordinated behaviour that emerges across society. Large-scale social simulations, or 'digital twins' of interacting social groups, help anticipate how communities might respond to top-down decisions prompted by events such as pandemics or climate change.

A related strand asks how to implement and validate ethical behaviour in autonomous systems: modelling social contexts, norms and conflicting values, prioritising goals, and simulating how individual behaviour shapes collective norms. This work is interdisciplinary, drawing on psychology, sociology and policy analysis.

**People:** [Vivek Nallur](people/vivek-nallur.md)

**Keywords:** Social Simulation, Machine Ethics, Normative Systems, Affective AI, Digital Twins


---

HTML version: ../research.html  
Site index for LLMs: ../llms.txt