Talking Buildings Adaptive Sensing Infrastructure towards Understanding Patterns of life
Talking Buildings: Adaptive Sensing Infrastructure towards Understanding Patterns of life across Heterogeneous Smart Buildings
Modern smart buildings are equipped with IoT sensors to facilitate efficient and effective maintains of buildings. These IoT sensors can be used to measure quite valuable aspects of buildings such as structural health, occupant behaviours, occupant health, and many more towards increasing functionality, comfort, safety, and reducing running costs. Even though much academic work has been done to generate these insights from sensor data, deploying them in the real world is quite challenging due to the simplistic assumption made within academic work. A more viable option is to buy very expensive off-the-shelf solutions from companies specialising in Buildings Management Systems (BMS) or Buildings AI solutions providers. The downside is that these solutions are often highly restrictive in terms of capabilities, extendability and adaptability. For example, we will be required to deploy their sensors exactly as prescribed and require a lot of manual labour to adapt them to new building types and layouts. Further, most of these BMS and AI solutions are designed to be used by domain experts (e.g., estate people who have specialised knowledge on energy standards, sustainability standards and so on).
In this project, we aim to address two key issues highlighted above. First, we will investigate how we could develop a semantic interoperability layer between IoT sensors and data analytics so the analytics could be adaptable for a given building’s configuration and layout. We aim to embed the domain knowledge into the system we are building so non-domain experts can use the system to understand better how the buildings are performing. To make the system more accessible, we aim to utilise conversational AI techniques to mediate the communication between the building and the non-experts. By doing this, we aim to give a voice to the buildings so they can communicate with humans in natural language and express how it feels. We envision a future that the buildings will be able to answer its performance-related questions (e.g., Building Research Establishment Environmental Assessment Method (BREEAM)) with the help of IoT sensors. This project has the following objectives:
- Conduct a literature review on the relationship between useful insights and what data types and analytics are required to generate such insights in the context of built environments.
- Extend (or combine) existing building ontologies to develop a semantic interoperability layer between IoT sensors and data analytics so the analytics could be adaptable for a given building’s configuration and layout, which can also handle the heterogeneity of IoT sensors.
- Develop a library of adaptable analytics that can generate insights using IoT sensor data while handling heterogeneity in terms of accuracy, reliability, sampling rates, etc.
- Implement a conversational AI-driven chatbot that can facilitate natural language communication between the non-experts and the building itself about its performance.
The Building Research Establishment (BRE) is a centre of building science in
the United Kingdom, owned by a charitable organisation, the BRE Trust. BRE
provides research, advice, training, testing, certification and standards for
public and private sector organisations in the UK and abroad.