High Growth AI Accelerator

In collaboration with Foster + Partners

Foster + Partners is a global studio for sustainable architecture, urbanism, engineering, and design, founded by Norman Foster in 1967. The practice is characterised by its integrated approach, bringing together specialists in architecture, engineering, interior design, and other disciplines to create innovative and holistic design solutions. With projects spanning the globe, the firm is committed to sustainability and creating high-performance, energy-efficient buildings and spaces that improve the quality of life and the built environment. 

The golden thread 

• How might we streamline data entry, improve collaboration, and ensure compliance across the diverse stakeholders in the AEC industry for safer, more efficient project delivery? 

The architecture, environment, and construction (AEC) industry involves diverse stakeholders, including architects, engineers, contractors, and facility managers, who rely on different tools and systems, resulting in fragmented data and communication gaps. Additionally, regulations such as the Building Safety Act 2022, require detailed documentation throughout a project’s lifecycle, especially for high-risk buildings.  

The golden thread concept aims to maintain a continuous digital record of information, ensuring the safety and compliance of high-risk buildings, tracking design intent and changes for effective lifecycle management. In the current landscape, the AEC sector struggles with manual data entry, necessitating input into multiple systems across projects with varied requirements. This redundancy increases the likelihood of errors, complicates collaboration, and negatively impacts compliance.  

This challenge seeks AI/ML proposals that can track, trace, and comply with project-specific requirements, and alleviate the data entry burden by: 

• Creating application-specific plugins (e.g., for Revit) for automatic logging and project-specific data collection. 

• Developing a unified system to automatically aggregate and tag data from different sources, ensuring a single source of truth for safety compliance. 

The proposed solution should not only improve project management but significantly enhance safety and compliance outcomes for higher-risk buildings. 

Automated architectural drawings 

• How might we create a secure, customisable tool that empowers architectural firms to automate drawings while preserving their unique styles and maintaining data privacy? 

Creating architectural drawings is a time-consuming and labour-intensive process, and automation can significantly reduce the effort required and increase efficiency. However, architectural firms typically have unique styles and design principles, and ensuring automated systems adapt to these specifications without compromising creativity presents a challenge. 

Maintaining consistency and accuracy in automated drawings is essential to ensure they meet industry standards and client expectations. Yet, many architects are hesitant to use AI/ML solutions from external providers due to concerns about data privacy, confidentiality, and intellectual property protection. 

This challenge seeks AI/ML solutions that architectural firms can independently use to align automated drawings with their unique style and design guidelines. By enabling in-house customisation and eliminating the need for third-party services that require sharing proprietary data, the solution should protect intellectual property and data privacy while enhancing creative outputs. 

In collaboration with Versarien

Versarien is committed to transforming the construction industry through advanced 3D construction printing, design and consultancy services, alongside the utilisation of graphene admixtures for eco-friendly construction. Our mission is straightforward: to deliver high-quality, efficient, and sustainable construction solutions. 

3D printing ideation 

• How might we streamline the process of generating 3D construction printing scripts to improve design efficiency and accessibility for designers? 

Versarien specialises in utilising 3D construction printing (3DCP) to produce modules of various shapes and sizes, using a printable structural mortar, built up in layers by pumping the material through a nozzle attached to robotic system, specifically a computer-controlled robot arm. One of the primary challenges in 3D Construction Printing lies in the time required to design customer-specific models and ensure their printability. Manually modelling these designs by physically manipulating geometry within a 3D design package consumes significant time and effort. It is often necessary to create multiple versions of the design, either to offer options or to create a collection of modules for larger projects. 

Parametric scripting, such as through Grasshopper, allows for the rapid generation of numerous geometric iterations by simply adjusting values. However, this approach is only feasible when the scripts already exist. Developing these scripts is itself a time-intensive process that demands a high level of expertise. Mastery of various Grasshopper modules, their functions, and their interoperability limits the number of individuals who can effectively use this technology, thereby restricting its application. 

This challenge seeks AI/ML solutions that can automate the generation of 3D Construction Printing scripts from natural language inputs, improving efficiency and accessibility. Future versions of this system could also integrate features like structural analysis to ensure that generated scripts produce printable geometries. 

Durability prediction 

• How might we improve the durability and performance of 3D printed structures by understanding key material characteristics and incorporating sensor data? 

The predictive models currently available for mortar are not specifically designed for 3D construction printing (3DCP), which imparts a specific set of characteristics to the material that are not well understood by the construction industry. For instance, the layering process introduces mechanical anisotropy in the structure, and the resulting increased porosity negatively impacts the mechanical properties and durability of the structures. Therefore, to consider the specific characteristics of this method, 3D printing construction should requires tailored tools. 

While it is possible to optimise the material’s mechanical properties (e.g.: by adjusting raw material quantities), there are limitations in addressing other critical factors such as porosity and water permeability, which significantly affect long-term durability. Additionally, there is a shortage of sensors capable of monitoring key factors like water ingress and pH levels, which are essential for predicting freeze-thaw degradation and carbonation, as most existing sensors only measure temperature and strength. 

The lack of durability data for 3D printed structures is a significant barrier to the widespread adoption of this technology. This challenge seeks AI/ML proposals that can develop tailored predictive models for 3D construction printing, incorporating live sensor data and aim to enhance the durability and performance of 3D printed structures. 

In collaboration with Buro Happold 

Buro Happold is a global consultancy of engineers, designers, and advisers dedicated to delivering innovative, value-driven solutions that benefit people, places, and the planet. With over 45 years of experience, they collaborate with leading architectural practices and organisations worldwide, including the United Nations, UNESCO, and C40 Cities, to address complex engineering challenges in the built environment. 

Design Interoperability 

• How might we integrate Digital Design workflows across diverse disciplines and project stages to enhance collaboration and efficiency? 

Buro Happold provides a full range of services, from project inception to asset handover and ongoing optimisation during the project lifecycle. To deliver these services efficiently, the company uses integrated digital design workflows supported by an open-source framework called the Buildings Habitats Object Model (BHoM).  

BHoM allows interoperability between 20+ design tools through 20 adapters, enabling better collaboration, faster optimisations, and higher-quality designs. However, Buro Happold’s internal catalogue includes over 1,000 tools, connecting only a fraction used across the diverse disciplines within the firm. This percentage becomes even smaller when accounting for workflows involving external collaborators in the broader project ecosystem. Creating adapters for these tools requires building new plug-ins and APIs each time, which is time-consuming and often not scalable for single projects or firms.  

This challenge seeks AI/ML solutions that can streamline and scale the process of connecting the workflows, enabling faster and more efficient integration of additional tools, disciplines, and collaborators into the BHoM framework. This could involve automating the development of adapters, analysing tool APIs to suggest mappings, or other AI-driven approaches. 

Visit bhom.xyz for more details about the framework. 

Industrialised Construction 

• How might we improve the efficiency of Industrialised Construction design by integrating manufacturing and assembly knowledge from the outset? 

In the early stages of project design, the knowledge of manufacturing supply chains, logistics, and assembly constraints is often unavailable or not well integrated into the design process. This lack of accessible and actionable information can lead to costly redesigns, delays, or inefficiencies making it challenging for design teams to fully optimise for Design for Manufacturing and Assembly (DFMA). While several approaches have been proposed to address this issue, they are often difficult to implement or not commercially viable. These approaches include: early integration of supply chain, designing around predefined “Construction Systems”, or process-oriented DFMA (e.g., CIH / IPA). 

This challenge seeks AI/ML solutions capable of transforming the design process by predicting manufacturing challenges, modeling supply chain constraints, and seamlessly integrating into existing workflows. Solutions could include AI-driven tools that provide real-time insights, simulate supply chain dynamics, or offer actionable guidance to enable smarter, faster, and more cost-effective decision-making from the very start of a project. 

An invitation to present your proposal to the built environment 

If the industry partner-led challenges do not align with your current project goals or focus area, we invite companies to present their ideas that can address pressing issues aligned with the construction sector an any of its subsectors. These may include, but are not limited to: Building Design and Architecture, Construction Materials and Technologies, Project Management and Construction Operations, Infrastructure Development and Maintenance, Building Information Modelling (BIM) and Digital Construction, and Environmental Sustainability. 

Your project must: 

• Demonstrate an innovative and ambitious idea that is technically feasible and scalable. 
• Utilise AI/ML or enable the use of AI/ML. 
• Demonstrate an increase in business productivity, contributing to the growth of the UK’s construction sector. 
• Address specific challenges faced by the industry. 
• Consider the potential impact of the proposed solution on the built environment, benefiting stakeholders such as contractors, architects, engineers, suppliers, and clients. 

The programme is inviting applications from UK-registered startups, scaleups and SMEs that:  

• Have have existing or new AI-enabled services or AI-integrated infrastructure solutions that can demonstrate to solve one of the challenges of the call 
• Have strong technical teams 
• Have available data and an immediate need for computation 

Applications will be assessed and scored equally against five criteria. All applications must have an AI/ML use case to take part in the programme but we encourage applications that wish to integrate any other advanced digital technology.

1. Relevance and feasibility
   • The applicant should demonstrate their solution can tackle the challenge selected and the company has the appropriate technical expertise to deliver the solution.
2. Business strategy
   • The applicant should be able to articulate the company’s business model that drives their AI/ML solution implementation and commercialisation.
3. Data and code readiness
   • The applicant should demonstrate that their company has the necessary data ready and has an implementation plan that requires immediate access to computational power.
4. Ethical impact
   • The applicant should exemplify a responsible use and understanding of the impact of their AI/ML solution, and a strong commitment to ethical AI practices.
5. Growth potential
   •  The applicant should be capable of identifying clear goals and demonstrating their solution has the potential to scale after the programme.

The scoring criteria will be assessed based on statements in the areas above. Each criterion will be scored on a range from 0 to 5. 0 being an “Unacceptable or No submission” score for each criterion and 5 being an “Excellent” score for each criterion. This scoring will be applied to all applications and will be equally weighted (20%). 

1. Applications judged and shortlisted
   • Applications will be initially screened for eligibility, followed by assessment based on selection criteria by the Digital Catapult team, resulting in the selection of a shortlist.
2. Interview day and selection
   • After the assessment of applications, companies will be notified of the status of their applications. Shortlisted applicants will be invited to an interview with Digital Catapult and the relevant Industry Challenge Owners. During the interview, applicants will have the opportunity to present their ideas and address any questions posed by the judges. Following the interviews, the panel will deliberate and select the final cohort.
3. Contracting
   • Successful applicants will be notified and provided with a standard Programme Agreement for review. These contracts are standard and not negotiable. We do try to ensure these contracts are fair and reasonable. Invitation to a programme kick-off will then follow provisionally on the completion of this agreement.

Application Process

1. Applicants check they meet the programme’s specific requirements.
2. Applicants fill out their application form through the platform Skipso. Applicants will need to complete the application form by 23:59 on 23 February 2025.

• Open call opens – 21st January 2025 
• Open call closes –  23rd February 2025 at 23:59   
• Q&A Sessions –  31 January, 7, 14, 21 February 2025 * 
• Notification of shortlisted projects and invitation to interview – 7th March 2025* 
• Interviews – w/c 17th March 2025 
• Notification of successful projects – 21 March 2025* 
• Contracting – 24 March – 4 April 2025  
• Programme Start Date – 7 April 2025
• Kick Off programme event – 9 April 2025  
• Programme end date – 11 July 2025  
• Final programme event – 10 July 2025 

*Dates may be subject to change