Human-Robot Teamwork Requirements Taxonomy
A NASA- and VeTSS DSTL-funded taxonomy of Human–Robot Teamwork requirements, grounded in international standards and validated across six domains.
Read more →
Dr Hazel Taylor
Research Scientist & AI Development Consultant
Trustworthy & Responsible AI · Regulated Environments
I’m a Research Scientist at the University of Manchester, where I lead NASA- and DSTL-funded research on Human-Robot Teamwork requirements and explainable AI for safety-critical, regulated environments, spanning nuclear, national security, and critical national infrastructure. I also work as an AI Development Consultant, building and shipping production AI systems, from RAG pipelines to human-in-the-loop automation, for clients in regulated sectors.
Projects
AI-Automated Invoice Extraction & Validation
A LangGraph pipeline that reads invoices with a self-hosted vision-language model, validates them against reconciliation rules, and routes exceptions to human review.
Read more →UK Company Filings RAG: Grounded Financial Q&A System
A well-tested prototype RAG system for querying UK company filings, engineered to eliminate hallucinated financial figures.
Read more →Human-Verified Dataset Builder
A human-in-the-loop pipeline that turns raw documents into clean, labeled datasets, reviewed through a local Streamlit UI and refined by feeding accepted corrections back into the model as few-shot examples.
Read more →Padel Shot Classifier
An end-to-end MLOps pipeline that detects and classifies padel shot types from match video.
Read more →Education
-
PhD Computer Science
Thesis: “Enhancing Human-Robot Interaction in Nuclear Environments: Explainability Requirements for Autonomous Systems.” Built and validated an explainable AI system used to justify autonomous robot deployment to nuclear regulators, partnering with RAIN (£112M UKRI/EPSRC programme) and nuclear operators.
-
MEng Robotics, Autonomous & Interactive Systems (1st Class Hons, Distinction)
MEng project: designed a cognitive architecture for scalable robot control in hazardous environments, published at IROS 2019. BEng project: built an ML-enhanced laparoscopy tool with real-time haptic feedback for surgeons. Deputy Principal’s Award for Academic Excellence, 2017/18 & 2018/19.
Research
As Principal Investigator on NASA- and VeTSS DSTL-funded programmes (the latter a competitively awarded £50k grant with a 7% acceptance rate), I lead development of a Human-Robot Teamwork requirements taxonomy validated across six domains, including a catalogue of 21 explainability requirement patterns formalised in NASA’s FRET tool and validated via NuSMV equivalence checking. I received my PhD from the University of Manchester for my work on “Enhancing Human-Robot Interaction in Nuclear Environments: Explainability Requirements for Autonomous Systems”, partnering with the £112M UKRI/EPSRC RAIN programme and nuclear industry stakeholders.
Research interests
Workshops
TAROS 2026 Workshop: Towards Effective Human–Robot Teamwork
Organiser of a workshop at TAROS 2026 on requirements and explainability for human–robot teams.
Visit workshop site →NASA Formal Methods (NFM) 2026 Workshop: Formal Requirements Engineering and Artificial Intelligence (RE×AI)
Co-organiser, with Marie Farrell, Divya Gopinath, and Anastasia Mavridou, of a workshop at NASA Formal Methods (NFM) 2026 on the synergy between formal requirements engineering and AI for assuring trustworthy learning-enabled systems.
Visit workshop site →iFM 2024: International Conference on Integrated Formal Methods
Organising Committee Member for the 19th International Conference on Integrated Formal Methods, hosted at the University of Manchester in November 2024.
Visit conference site →Publications
-
A Taxonomy of Human-Robot Teamwork Requirements
-
Engineering Reliable Autonomous Systems: Challenges and Solutions
-
All Dressed Up: Requirements for Human-Robot Teamwork
-
Eliciting Explainability Requirements for Safety-Critical Systems: A Nuclear Case Study
-
Explainability Pattern Specification for Human-Robot Teamwork
-
Towards A Catalogue of Requirement Patterns for Space Robotic Missions
🏆 Best Paper Award
-
What to Explain and How? The Challenges and Future of Using Large Language Models to Generate Explanations for Lawyers in Autonomous Car Accidents
-
Enhancing Human-Robot Interaction in Nuclear Environments: Explainability Requirements for Autonomous Systems
-
Should AI Systems in Nuclear Facilities Explain Decisions the Way Humans Do? An Interview Study
-
An Abstract Architecture for Explainable Autonomy in Hazardous Environments
-
Robots in the Danger Zone: Exploring Public Perception through Engagement
-
Introducing a Scalable and Modular Control Framework for Low-Cost Monocular Robots in Hazardous Environments
- Full list with citation counts on my Google Scholar profile.
CV
For a full, up-to-date CV covering education, publications, and funding, please contact me via email or connect on LinkedIn.
Contact
Eligible for UK security vetting.