Hacking the City to improve existing buildings

I expected that participating in the FastTrack CityHack in Hong Kong late last year was going to be intense, what I had not predicted was how it would expand my knowledge and inventiveness as an engineer.

As a structural engineer, my work involves a significant amount of designing beams and columns. CityHack however, took me out of that professional comfort zone to consider how Artificial Intelligence and smart glazing could be used to reduce energy for cooling and improve human comfort inside buildings.

We were randomly assigned to teams of five people from different companies and with varying professional and academic backgrounds, then tasked with developing a scalable, smart solution to improve the performance of existing buildings.

This is a challenge generally across our built environment. For example, there are smart glazing panels on the market that can respond to daylight and external conditions to reduce heat gain and glare and minimise energy use for cooling. However, from my experience, replacing all the glazing in a building to retrofit smart panels can be a very expensive and challenging task.

The alternative solution for existing buildings is smart window films that can be applied over existing glazing to provide a degree of light-responsive automated shading and reduced heat gain.

The idea my team and I proposed for the competition was to develop a system that adds advanced coordination capabilities, including machine learning functionality that responds to the habits of occupants to automate the response, not only in reaction to weather conditions but to people’s habits and needs.

Our team collaborated to design a surface that connects multiple units of glazing and is able to responsively change the opacity of the glazing to block out or filter daylight if required, or allow full daylight penetration in the right conditions.

The AI within the system incorporates machine learning that accumulates data on local weather conditions, so the response of the system is more tailored. This goes beyond basic seasonal changes such as length of day, average temperature and so on to also include real-time responsiveness to such occurrences as a cloud covering the sun. This enables a microclimate level of adjustment.

It is not only external conditions the system will respond to. The use of sensors in the system means it can also adjust light penetration depending on whether a person is simply walking through a space or if they are standing in front of the window. We also added a layer in the AI platform which means individuals can specify whether they prefer a greater or lesser level of daylight in their workspace.

This is a form of granular data we do not yet have easy access to, around human preferences and comfort and what is optimal for different individuals in terms of daylight levels. Our system can begin to collect that data, and this will have value for decision-making in other buildings and projects, particularly for smart buildings whether they are new or retrofitted.

Our system can also be connected into an existing BMS system, so there is minimal need for new digital hardware or wiring.

We named the system “Green Glaze” as it will also help reduce the carbon footprint and operating costs of buildings. Currently, in Hong Kong, energy for cooling is a substantial part of overall energy consumption, so a solution that reduces that has a very meaningful impact for not only building owners and occupants but also for our national energy system. Our concept was the 2nd runner-up and won a HK$10,000 prize, and we have now moved as a team into looking into commercialisation.

To be honest, at the start of the competition I was just a little bit sceptical. The structures and buildings I have been working on at Cundall are – I thought – already smart enough. I did not think we need to cut the cake from every side.

However, we do work as one cross-disciplinary team, so I am always learning from our building services and sustainability experts. I had many conversations with them as I was developing the idea around what the impact would be of smart glazing and what are some of the needs it can address.

Since the CityHack competition, I now see how much room there is for improvement in our current building design and building systems, and this also helps me in my own work to see ways to improve the sustainability of outcomes.

As building and engineering practitioners, there is so much we can do to improve the buildings we already have. I have a new appreciation for the possibilities AI and machine learning offer and am actively looking at ways they can be implemented to improve the solutions I help create.