An Environmental ‘App’ for Architects

An Environmental ‘App’ for Architects: Utilising Artificial Neural Networks and Real World Data to Predict Operational Energy Consumption of School Buildings Based on Early Design and Briefing Decisions 

By Greig Paterson

Supervisors:

Professor Dejan Mumovic 

Dr Payel Das 

Industrial Supervisor/Sponsor:

AHR (formerly Aedas Architects)

2010-2015

Aim:

Create the prototype of a user-friendly early stage design tool that predicts operational energy consumption of school buildings based on the training of artificial neural networks with real world data.

Overview:

It has been argued that traditional building simulation methods can be a slow process, which often fails to integrate into the decision making process of non-technical designers, such as architects, at the early design stages. Furthermore, research, such as that carried out by CarbonBuzz, highlights the fact that the actual energy consumption of buildings regularly exceed design estimates, often by more than double.

In view of this, a user-friendly design tool is being developed in the form of a simple ‘app’, which predicts building performance in real-time as early design and briefing parameters are altered interactively. As a demonstrative case, the research focuses on school design in England. Artificial neural networks (ANNs), which are a subset of artificial intelligence, have been trained to predict the heating and electricity energy consumption of school designs by linking measured energy consumption data from the building stock to a range of early design and briefing parameters. 

The artificial neural networks have learned through observations of real world data – a technique that may help reduce the performance gap between predicted and actual energy consumption.