Projects¶
This page brings together selected funded projects, research collaborations, and open-source software connected to dependable, safe, and explainable AI.
Project Directory¶
The projects below are based on the official Repository@Hull Worktribe records for Dr Koorosh Aslansefat.
| Project | Dates | Role | Funder / Partner | Focus |
|---|---|---|---|---|
| AKT TRUST-LLM | 2024 | PI | Innovate UK / Walton & Co | Responsible and trustworthy LLMs for planning-law institutional memory. |
| KTP: Connexin | 2023-2025 | Co-I | Innovate UK / Connexin | Speech-to-text, sensor networks, smart speakers, and adult social care. |
| iCASE PhD Studentship with QinetiQ | 2024-2027 | Co-I / supervisor | EPSRC / Naimuri, a QinetiQ company | LLM assurance, hallucination detection, and model correctness. |
| iCASE with QinetiQ via Newcastle University | 2024-2027 | Co-I / supervisor | EPSRC / QinetiQ / Newcastle University | Knowledge-enhanced LLM assurance. |
| Digital Twin for Offshore Wind Turbine Gearing | 2024-2025 | Co-I | University of Hull / IIT Madras | Physics-informed digital twins for fault diagnosis and prognosis. |
| Transforming Urban Health | 2025 | Co-I | University of Hull / IIT Madras | IoT indoor-air-quality monitoring and citizen-centred health interventions. |
| Vital Knowledge Exchange Visits | 2024 | PI | University of Hull / IIT Madras | Responsible AI knowledge exchange and collaboration building. |
| Alan Turing Institute Post-Doctoral Enrichment Award | 2022 | Awardee | Alan Turing Institute | SafeML and machine-learning safety monitoring. |
Grants and Collaborations¶
The full funding portfolio is listed on the Funding page. Selected collaborations include:
| Role | Project | Partner or funder |
|---|---|---|
| Academic Lead | Trustworthy AI Agent for Incident Management | Innovate UK and Veracity Healthcare |
| PI | TRUST-LLM: Responsible Use of Safe and Trustworthy LLM for Planning Law | Innovate UK and Walton & Co |
| PI | OCTOPUS: Mobile Robotic Systems and AI for Monitoring Ports and Estuary Waters | International collaboration with France, Spain, and South Korea |
| PI | Vital Knowledge Exchange on Responsible AI Projects | UoH-IIT-M and IIT Madras |
| Co-I | LLM Assurance Framework through SafeML | EPSRC, QinetiQ, and Newcastle University |
| Co-I | Responsible AI for Care Chatbots | Innovate UK and Connexin |
| Co-I | National Edge AI Hub for Real Data | EPSRC and 50+ industry partners |
| Co-I | Supergen Offshore Renewable Energy Impact Hub | EPSRC and 44+ industry partners |
| Research Associate | SESAME: Secure and Safe Multi-Robot Systems | EU H2020 |
| Research Assistant | GO0D MAN: Agent-Oriented Zero Defect Manufacturing | EU H2020 |
Open Source¶
-
SafeML
Statistical safety monitoring for machine-learning classifiers. SafeML has been used for uncertainty quantification in safety-critical applications and was recommended in German Industry Standard DIN SPEC 92005.
-
XWhy / SMILE
Robust model-agnostic explanation methods for safety-critical machine learning, including Statistical Model-agnostic Interpretability with Local Explanations.
-
SafeDrones
Runtime reliability evaluation for UAVs and multi-drone systems using executable digital twins and safety-aware coordination methods.
-
SafeLLM
Domain-specific safety monitoring for large language models, with applications in wind maintenance and industrial decision support.