The XR5.0 consortium continues to make strong progress in the development of Pilot 3, which explores how Artificial Intelligence (AI) and Extended Reality (XR) can support water utility operators in inspection, maintenance, and training activities. The pilot aims to improve situational awareness by combining AI-driven analytics with intuitive visualisation and natural interaction technologies, enabling operators to better understand underground infrastructure and make more informed decisions in the field.
Pilot 3 focuses on three complementary use cases addressing key operational challenges faced by water utilities. The first supports maintenance activities by visualising underground pipe networks together with real-time sensor information to facilitate the identification of potential leakages. The second enhances CCTV pipe inspection workflows by presenting AI-detected anomalies alongside explainable AI outputs, allowing operators to better understand the reasoning behind automated detections. The third use case provides an AI-assisted training and operational support environment through an intelligent conversational assistant that enables natural interaction using both text and voice.
Over the past development period, the CYENS team has implemented a broad range of functionalities that bring these use cases closer to operational deployment. These include real-world geolocation and underground pipe visualisation, integration of IoT sensor data, AI anomaly visualisation, conversational AI services, speech-to-text and text-to-speech capabilities, modular backend communication, and a flexible user interface designed to present contextual information when and where it is needed.
A key milestone has been the evolution of the platform through continuous user feedback. Initial prototypes were developed for Magic Leap augmented reality devices, allowing operators to visualise infrastructure directly within their surroundings. User evaluations, however, highlighted several practical considerations related to navigation, information density, readability, and safety during field operations. These findings guided a significant redesign of the platform, including support for mobile devices, simplified navigation, improved visualisations, enhanced AI explanation panels, redesigned information windows, and integrated chatbot functionality. The result is a more accessible and user-friendly solution that better reflects the needs of end users.
The pilot has also addressed several technical challenges associated with deploying XR applications in real-world outdoor environments. Development efforts focused on improving GPS positioning, reducing compass drift, managing heterogeneous backend services, handling network latency, and ensuring reliable speech interaction. A modular software architecture and reusable components have been adopted to facilitate future scalability and integration with additional services.
Development has now reached an advanced stage, with the maintenance support use case nearing completion, AI-assisted inspection capabilities entering their final validation phase, and the training environment continuing to mature through interface refinements and additional testing. Upcoming work will focus on further optimisation, expanded user evaluations, single sign-on integration, and preparation for the final pilot demonstrations.
By integrating AI, explainable decision support, and Extended Reality into everyday operational workflows, Pilot 3 demonstrates the potential of human-centric digital technologies to support more efficient infrastructure management while keeping human expertise at the centre of decision-making. As XR5.0 progresses towards its final demonstrations, the pilot continues to contribute valuable insights into the practical adoption of Industry 5.0 technologies within the water sector.