Whole-building performance has emerged in response to this need, recognising the interconnectedness of different objectives and how improvements in specific areas can have ripple effects on others. But how can multiple objectives be balanced?

Daisy, our AI-powered, parametric design tool, can achieve just this. It is underpinned by a co-design philosophy, which utilises systems-level thinking and digital solutions to bring architects, engineers and analysts together at the early-stage concept phase to address the complexities of whole-building performance. 

It runs thousands of AI-driven simulations using insights from multidisciplinary experts and client goals. It finds the best balance between factors like daylight, thermal comfort, energy use and embodied carbon. This helps project teams to make better decisions and improves the overall performance of a building.

Benefits of this approach include:

• Improved daylight: In areas with poor daylight, specific upgrades to the building's exterior can be suggested to improve light levels without significantly increasing energy use.
  
  
• Reduced overheating risk: In areas prone to overheating, passive design strategies like adjusting shading and window sizes can be proposed to reduce heat without affecting daylight. This removes the need for extra cooling systems.
  
  
• Optimised energy efficiency: This can uncover optimal glazing ratios and the window spacing needed to maximise energy efficiency.

Daisy draws inspiration from regenerative design principles, emphasising shifts such as moving from opinions to evidence-based problem-solving and from mechanistic to systems thinking. This sees systems as connected and focuses on how they interact, rather than just looking at individual parts. It's about finding solutions that balance different factors to optimise whole-building performance.

This approach is underpinned by the below four key principles – multidisciplinary collaboration, early-stage asset simulation, mass computing and AI for simulation, and design trade-off characterisations.

1. Multidisciplinary collaboration

Traditional design processes involve architects handing over work to engineers in a one-way manner, leading to compartmentalisation and fragmentation. This method doesn't address the interconnected nature of building performance. Optimising one aspect late in the design, like daylight, often compromises others, such as energy use or embodied carbon.

Our co-design tool, Daisy, promotes collaboration by defining whole-building performance objectives like energy efficiency, embodied carbon, daylighting and thermal comfort.

2. Early-stage asset simulation

In many projects, engineers aren't involved early enough when there's the most flexibility in design but the least knowledge. This gap makes it hard to make complex decisions needed for overall building performance from the start. To fix this, services should be adjusted to evaluate the entire building's performance early in the design process. By thoroughly examining how buildings behave as a whole, new insights can be discovered and misunderstandings corrected early on to prevent costly changes during the later stages of the project.

3. Mass computing and AI for simulation

Daisy uses computational simulation and multi-objective optimisation to provide essential data for early-stage design decisions. Computational methods like generative design, automated simulations, parallel computation and machine learning help assess complex designs. 

Multi-objective optimisation generates thousands of valid design options, each evaluated across multiple performance objectives. This process identifies efficient designs that achieve optimal trade-offs, ensuring improvements in one objective result in minimal compromises in others.

4. Design trade-off characterisations

After simulating a range of designs, the data can be used to measure and describe the trade-offs between KPMs and the relationships between design variables.