Research Programme

Talking Buildings: Adaptive Sensing Infrastructure

(2021-2025)

Smart Buildings Conversational AI

Internet of Things (IoT) Infrastructure / Systems (IS) Human Computer Interaction (HCI) Sustainability (SU)

Project Overview

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

Modern smart buildings are equipped with IoT sensors to facilitate efficient and effective maintenance of buildings. These sensors can measure structural health, occupant behaviours, and occupant health. However, existing commercial Building Management Systems often lack the flexibility to handle heterogeneous sensor deployments and provide accessible interfaces for non-expert users.

This project addresses these gaps by developing adaptive sensing infrastructure that enables buildings to communicate their operational status through conversational AI. The research focuses on creating semantic interoperability layers and natural language interfaces that allow facility managers, occupants, and researchers to query building performance data without requiring technical expertise.

Objectives

Conduct a comprehensive literature review connecting useful insights to required data types in built environments
Build a semantic interoperability layer extending existing building ontologies to handle IoT sensor heterogeneity across different manufacturers and protocols
Develop adaptable analytics that handle variations in sensor accuracy, reliability, and sampling rates
Create a conversational AI chatbot enabling natural language dialogue about building performance metrics and conditions