Brains for Buildings
How to design a data-driven interface that supports the user in understanding the environmental decisions made by the energy management system and supports the user in making his environmental decisions on a decentralised level while considering a healthy indoor climate, energy efciency, and energy flexibility?

Main Problem
In most modern utility buildings 10-30 per cent of the energy is being wasted by malfunctioning installations and unexpected user behaviour. Also, the indoor environment is inadequate, and the operational costs are high. Smart meters, facility management systems and the Internet of Things offer the possibility to gather extensive amounts of data.
This research is part of the Brains for Building’s Energy Systems (B4B). Brains for Building’s Energy Systems is a multi-year, multi-stakeholder project focused on developing methods to harness big data from smart meters, building management systems and the Internet of Things devices, to reduce energy consumption, increase comfort, respond flexibly to user behaviour and local energy supply and demand, and save on installation maintenance costs. This will be done through the development of faster and more efficient Machine Learning and Artificial Intelligence models and algorithms. The project is geared to existing utility buildings such as commercial and institutional buildings.
By implementing Smart Systems like Machine Learning and Artificial Intelligence to the Energy Management System (EMS) of utility buildings, the captured real-time senor data can be analysed and used to reduce energy consumption while optimizing the user’s comfort.
An Artifcial Intelligent EMS can ensure that the rooms in the building are optimized for the user. In an optimal situation, the AI is always correct about the user’s preferences and the user will always be comfortable. Thus, he will not have to arrange anything himself. However, it is almost impossible to create a system that adapts to everyone's preferences and will be 100% correct about every user’s preferences. For the system to work as well as possible, the user must be able to talk directly to the AI system to steer the system in the right direction. For example, if the temperature set by the AI is not at the correct level, the user should be able to correct the AI so that it will set the correct temperature next time. The research is about how to make a design that could contribute to the communication between the AI-EMS and the user.
Positioning
As a designer, I focus primarily on research and design through data. I’ve always had an interest in data research. The focus on research and design through data started when I participated in the minor in Data Science at Rotterdam University of Applied Sciences. This minor provided me with the skills to do research through data like data engineering, data visualisation and machine learning. It also thought me how to use privacy techniques to create an ethically responsible data design. These skills helped me significantly during the research phase. I used them to derive answers and insights from the acquired real live data.

Research Results and Solution
The research showed that comfort can be divided into three main categories. These categories are visual, social and thermal comfort. To estimate the overall comfort of a user, it is most important to consider these three categories. Through user tests and data visualisation it is proven that adjustments in one of these three categories can significantly affect the user’s overall comfort.
Comfort can be used to communicate with the EMS system. When the user can provide their feedback on their overall feedback, the AI could use this data and the sensor data to predict the users comfort.
The final design is the Brains for Buildings feedback application. This mobile application ensures that the user has insight into the choices of the EMS system utilizing an easy interface. He can see the temperature, visual variables and social variables. Thereby, the user can get an insight in the AI’s decisionmaking. The user can then provide feedback about the choices of the EMS system using a simple interface. Are you very comfortable, or not at all? This allows the system to adapt itself to your preferences.
The application is part of a smart network where the sensor data of the building is distributed. Also, the AI is part of this smart network. Therefore, the AI can easily combine the sensor data with the comfort feedback to make a prediction.
All combined, this makes for an easy interface for the user to acquire insights about the AI-EMS and simuntaneously provides the user with the ability to give their feedback on the AI-EMS’s choices.