This project creates low-power sensing systems using spiking neural networks to detect threats in public spaces.

Coordinates UK research on edge AI resilience, focusing on cyber-disturbances, data quality, and time-critical applications.

This project develops a multi-model AI system to accurately identify and count wildlife in camera trap images.

This project develops tangible interfaces to assist young people with neurodiversity to manage online harms.

This project develop an augmented reality-based Smart City demonstrator using Lego towards public engagement

Cardiff’s PETRAS regional showcase brought researchers, industry, and policymakers together to share “Connected Spaces” projects and identify future collaborations.

Builds an open-smartwatch prototype that guides users through indoor spaces using BLE beacons and onboard navigation algorithms.

Prototypes a Raspberry Pi-powered hexapod that combines guidance lighting, computer vision, TinyML, and emergency communications to support ageing users.

This project aims to integrate hardware sensors and drone images to develop a scalable forest health index.

Designing a smart building platform that uses mobile IoT sensors to reduce deployment complexity while maintaining comfort, sustainability, and wellbeing insights.

Provides an isolated lab of 170+ smart home devices connected through openHAB for experimentation, data collection, and cyber-physical anomaly detection research.

Builds an AI-assisted tool that helps novice software engineers learn privacy-by-design practices and legal compliance through guided design activities.

Examines whether microcontrollers and TinyML can detect aggressive door slamming as an early warning signal for domestic violence.

Uses LSTM autoencoders with surrogate models and SHAP explanations to interpret anomaly detection decisions on IoT sensor data.

Commercialises the CASPER cyber-physical anomaly detection work into a product proposition through the UK CyberASAP accelerator.

Co-creating an Edge Analytics course with partners in the UK and India to equip engineers with skills in machine learning and IoT for edge devices while building long-term academic collaboration.

Integrates CASPER anomaly detection research with Thales UK’s autonomous logistics platform to support resilient decision making during UGV missions.

Builds a linked-data observatory and analytics platform to support forest conservation, community engagement, and scientific research.

Develops guidance, datasets, and tooling to evaluate IoT anomaly detection techniques across heterogeneous smart home testbeds.

Builds semantic interoperability and conversational interfaces so non-experts can query building performance data across heterogeneous smart spaces.

Uses knowledge-based modelling and the MAPE-K loop to orchestrate self-adaptive cyber-physical security for smart homes without relying on the cloud.

Investigates IoT and data-driven approaches to enable circular supply chains in construction, helping the built environment meet net-zero goals.

Investigates how low-cost cameras can complement sensor networks to detect anomalies in smart environments.

Develops FedBio-IoT, a federated self-configuring IoT architecture using nature-inspired algorithms for runtime configuration and context-aware adaptation.

Combines pre-trained vision models with edge processing to flag unusual activity for smart city deployments in farms, castles, car parks, and bus stops.

Explores how to add layers of resilience to smart homes and offices by complementing traditional systems with independent sensing and analytics.

Deploying IoT sensing within Cardiff University’s Abacws building to characterise study-space usage and guide collaborative service design decisions.

Making linked data observatories accessible to bioscience researchers by blending graphical and conversational interfaces backed by large language models.

Develops distributed analytics that can move between edge, fog, and cloud nodes so hygiene services operate reliably without constant connectivity.

Designs resilient, low-power communications for monitoring traps, poaching activity, and sensors across dense jungle terrain in the Lower Kinabatangan Wildlife Sanctuary.

Proposes a secondary IoT sensor layer that monitors physical signals in buildings to uncover cyber attacks that evade traditional network monitoring.

Co-designs privacy tooling and conversational assistants that help older adults manage smart-home data while maintaining independence.

Combines network traffic analysis with independent sensor observations to detect cyber-physical anomalies in smart homes using DIY IoT nodes.

Trials immersive tourism and farming security services across rural Monmouthshire and Blaenau Gwent, leveraging 5G, edge analytics, and cyber security to grow the local economy.

Builds a study platform to explore how people value different IoT data types and which organisations they trust enough to trade with.

Extends the PizzaBox platform to help primary school children learn about balanced diets through playful, object-driven food ordering experiences.

Building a linked-data Forest Observatory that unifies heterogeneous bioscience and environmental datasets to help predict poaching activity in Sabah, Malaysia.

Builds CASPER, a context-aware anomaly detection platform that observes industrial robots with independent IoT sensors to spot attacks that evade traditional network monitoring.

Designs interactive technologies that encourage office occupants to adapt to mild discomfort, reducing energy consumption while maintaining wellbeing in built environments.

Demonstrates how semantic web technologies running on openHAB can wrangle personal IoT data at the edge, reducing unnecessary data transfers for smart-home data science.

Uses proximity between family members to adapt smart-home permissions, ensuring children access sensitive content only when guardians are nearby.

Generates realistic smart-home activity data by scripting scenarios through a GUI and publishing them to openHAB via MQTT so researchers can prototype analytics and interfaces without installing hardware.

Designing tangible interfaces such as PrivacyCube and PriviFy to help households understand and manage how smart-home devices handle their data.

Designing resilient sensing and communications infrastructure for Sabah’s Lower Kinabatangan Wildlife Sanctuary so conservation teams can gather data without relying on constant connectivity.

Explores how cosplay experiences change when audiences remotely control costume behaviours using connected hardware, demonstrated through the IoT Skullfort build.

Uses the open-source OLYMPUS platform to help learners analyse commercial IoT products together, surfacing design decisions and encouraging deeper peer discussion about connected-device ecosystems.

Evaluates Bluetooth Low Energy beacons and concealed receivers as a low-cost method to trace poacher vehicle movements in remote jungle terrain.

Deploying a low-cost SMS-based tracking system to understand poacher movements and disrupt the wider wildlife crime supply chain.

Prototyped PizzaBox, a tangible pizza-ordering interface that explores playful, healthy-by-design food journeys and real-time connected services through 3D-printed interactions.

Equipping IoT developers with reusable privacy-preserving components and gamified tooling so applications comply with global regulations by design.

Analysing CAN bus traffic and enhancing controller design to detect and prevent cyberattacks against connected vehicles.

Building semantic, on-demand data offerings for smart city marketplaces so consumers buy exactly the IoT data they need while reducing bandwidth and pricing friction.

Protecting critical infrastructure by using explainable AI at the edge to analyse data flows and detect vulnerabilities across industrial IoT systems.

Establishing a reconfigurable edge-computing testbed of Raspberry Pi and Jetson devices to explore latency-sensitive analytics, communications, and cybersecurity scenarios across campus spaces.

Investigates optimal placement and scheduling of service function chains (SFCs) across multi-cloud environments under security constraints.

Builds Privacy Captain, a knowledge-based AI assistant that annotates IoT designs with privacy patterns to reduce breakdowns in software design processes.

Develops interactive tools (Privacy Parrot) that bring privacy-preserving techniques into early IoT software design, supporting collaborative, privacy-aware workflows.

Develops modular IoT training programmes for diverse learner groups—from kids to makers—covering long-lived principles alongside rapidly evolving tools.

Benchmarks multi-criteria decision-making techniques (SAW, TOPSIS, VIKOR) for discovering IoT resources in the Cloud of Things, evaluating capacity and diversity trade-offs under varying user constraints.

Develops privacy-by-design guidelines and assessment methodology to help IoT software engineers incorporate privacy awareness during application design.

Builds the Sensing-as-a-Service model to enable data trading between IoT data owners and consumers, unlocking siloed datasets for broader value.

Explores fog and edge computing architectures for IoT, evaluating distributed analytics platforms and their applications.

Demonstrated seamful design techniques for BLE-based indoor experiences through the Ghost Detector museum game.

Surveyed more than 100 commercial IoT products to map technologies, applications, and gaps within the emerging sensing-as-a-service market.

Explores how natural-language sticky notes can act as end-user programming for smart homes, bridging human-written reminders and IoT automation.

OpenIoT delivers an open-source IoT middleware that abstracts heterogeneous sensors and services, enabling sensing-as-a-service deployments across cloud environments.