‍The current wave of innovation in architecture and engineering software is long overdue and accelerating at a rapid pace. But how do software advancements help us to decarbonise the global building stock? This insight outlines why design automation will be an essential enabler for the real estate sector to (finally) adopt green building principles at scale.

Green design premiums are a problem

• Design is the most important factor in determining GHG emissions over a building’s lifetime. By the time the construction process begins, the majority of decisions affecting a building’s lifetime emissions are locked in. As client preferences shift towards the total cost of ownership, design automation will empower architects to have greater agency to prove the value of their designs, accelerating green construction in the mass market.
• As buildings have grown in complexity, technology has been used as a tool to keep pace rather than to innovate, to digitise previously manual processes rather than to automate core aspects of the design workflow. Complex building regulations, limited access to product data and a scattered ecosystem of green building products make it difficult for architects to maintain code compliance and avoid greenwashing claims. Currently, the industry relies on a range of consultant experts, who are not only expensive but in short supply.
• Experts are typically engaged on bespoke projects, where a clear climate mandate from the developer prices in any additional consultancy costs or ensures a specialist architectural firm is engaged from the outset. In the absence of larger budgets, conventional value engineering practices in the mass market see architects’ initial product specifications swapped out for cheaper alternatives with often much worse environmental performance.
• Green materials can be standardised through prefabrication systems. This significantly de-risks application for generalist architects, reducing the need for specialised skillsets and minimising the risk of value engineering. However, designing with prefab brings other challenges. While architects design a building’s structural system pre-permitting, prefab system specifications traditionally occur post-permitting. This means that architects working on a prefab concept design are forced to build in large tolerances to account for the unknown variation in system specifications. In turn, significant inefficiencies are created later down the line, as designs often have to be painstakingly reworked once the prefab system is selected.
• Integrating performance analyses, embodied carbon data and prefabricated design system specifications within architectural software will empower architects to have greater agency to prove the value of their designs, reduce the reliance on specialised skillsets, and improve the efficiency of architecture firms.

Overcoming AEC incumbents

• The architectural software market is ripe for disruption, with rapidly rising software costs - 70% from 2015-2019 and 15% in the year to 2021 - fuelling outrage across the profession towards AEC software giant Autodesk. Architects’ frustrations are certainly justified, given increased costs appear to be met with declining productivity. With gross margins often as low as 9%, nearly two thirds of an architecture firm’s operational costs are spent on labour.
• Making a mark in architectural software is not without its challenges. A highly concentrated ownership structure (Autodesk holds 90% market share) is met with a customer need for a diverse set of software solutions on any given project. This requirement is due to the varying complexity of project design and a generational gap in technical skillsets among architects.
• A disparity in technical skillsets and complexity of learning new design software makes it impractical to train entire generalist architectural teams in the ‘latest’ solution. Among the most technical users are computational architects – typically BIM consultants in larger firms – who use complex visual programming tools such as Grasshopper to code custom shortcuts or add-ons for the wider firm. At the other end of the spectrum sit inhouse architects at housebuilders or prefab manufacturers, who use architectural software to produce near identical building designs. For this reason, documentation can often be generated through predefined Autodesk Revit blocks alone.
• Contractual requirements to use Autodesk Revit can leave architects little agency to implement directly competing products. For this reason, plugging into one gap in the existing workflow today tends to provide industry disruptors with the best entry point (and ample market opportunity) from which to scale automation throughout the wider design process.

Design automation

i. Improving interoperability

• Poor interoperability between existing software products creates tedious rework to manually update models one by one. Given up to 70% of architectural work is carried out on a fixed fee basis, rework is often at the architect’s expense. Long hours, tedious workloads and unpaid overtime contribute to an acute shortage of skilled workers in AEC, which is only exacerbated by a growing demand in green material expertise. To date, the industry has relied on open source collaborations between incumbents to improve interoperability. For example, Rhino.Inside - a collaboration between Autodesk’s Revit and McNeil’s concept design tool Rhino - enables architects to design in Rhino from within the Revit platform.
• New platforms are providing interoperability across the wider AEC software ecosystem. For example, Speckle’s inter-AEC software connectors enable updates from one design model to be reflected in another in real time. Integrations extend across a number of AEC programmes and gaming engines, including Unity and Unreal. For architects using Speckle, the platform significantly reduces manual design rework, enables inter-party tracked changes, and reduces the frequency of clashes between architectural, MEP and structural inputs.

ii. Reimagining concept design

• Recent years have seen environmental performance simulation become integrated within concept design software. This provides a dynamic, near real time insight into a proposed design’s future lifecycle impact. For example, browser-based tools such as Cove.tool enable architects to upload a design model and quickly receive detailed performance analysis data across carbon, daylighting, wind, water, density and cost.
• We have seen two emerging approaches. On the one hand, performance analyses are integrated within generative design software, with AI-generated massing and layouts optimised for environmental performance. In practice, the overly simplified user interfaces and limited user controls that these solutions provide sit at odds with the iterative and creative design workflow – aiming to standardise and replace the most custom aspect of building design.
• Other approaches – including Generate – augment architectural workflows, using AI to enable creativity assistance. The greater flexibility provided empowers architects to tune the degree of automation and consider a range of carbon and cost data points in real time. Meanwhile, Snaptrude’s cloud based concept design tool is as easy to navigate as SketchUp, while incorporating core aspects of Rhino and Revit and plugins for performance analysis tools. The goal is to create a platform with all the bells and whistles of modern software tools (collaboration, cloud-native, usability, etc.) with the power and flexibility of current design tools.

iii. Automating construction documentation

While the concept design stage comprises the most creative aspect of the architectural workflow, detailed and technical design stages represent the bulk of an architect’s fee (45-75%) and project time (70%) (A/O analysis). When it comes to prefabricated systems, multifamily, educational and healthcare buildings, unit-based designs make it possible to autonomously optimise technical design and engineering workflows.

• For example, emerging software platform SWAPP automates the tedious process of transforming an architectural design into a fully constructable building. By ingesting a number of architectural models and other datasets, the platform uses AI to optimise for custom architectural constraints, generating a BIM model and associated construction documentation with near complete accuracy.
• Automated design documentation extends beyond architectural workflows. Platforms - such as Augmenta - automate MEP and structural design. These solutions allow users to not only identify clashes between AEC datasets, but automatically resolve clashes through predetermined hierarchies. This is no easy feat given MEP optimisation alone relies on a range of unique parameters, including engineering calculations, product specifications, energy optimisation based on placement and specifications, and maintenance considerations.
• Striking the right balance between useability and flexibility will be key to realising mass market adoption of architectural automation software. This is due to the wide range in technical skillsets between architects. Platforms - such as Hypar - target a firm’s computational architects, for whom flexibility and platform extensibility is key. A small but growing market, computational design will prove critical for enabling architectural automation in the long term. Other platforms - such as SWAPP - combine a highly technical infrastructure with the ability to serve a non-technical audience (generalist architects) today. This provides immediate value for a large share of the current market, while retaining the option to open up the platform to computational architects later down the line.

iv. Enabling manufacturer integrations

Long term, automated design documentation will extend to the onboarding of structural and façade system manufacturers. This will enable generalist architects to:

• Quickly and accurately create buildable designs using green material prefabricated structures, such as mass timber
• Quickly and accurately optimise MEP and structural engineering with the chosen structural system
• Extract highly accurate pricing and carbon estimates in real time

At A/O we are committed to backing founders looking to digitise and decarbonise the built world. If you’re a startup working to transform the architectural workflow, please get in touch as we’d love to meet you!

Startups mentioned: Speckle, Snaptrude, Augmenta, SWAPP, Hypar, Cove.tool, Generate

‍