First-of-its-kind MSc teaches skills in generative design and design optioneering, making new AI tools for architecture possible
The UK’s first master’s degree to apply AI to engineering and design will explore potential architectural applications for design data analysis and generative design.
Based in Bath University’s Department of Architecture and Civil Engineering, and open to architecture post-graduates among others, the MSc in artificial intelligence for engineering and design will open its doors in October.
The multi-disciplinary course spans four univeristy departments in the Faculty of Engineering & Design and aims to teach how engineering and design knowledge, combined with data science and machine learning, can be applied to complex engineering problems.
Students will learn the key problem-solving and technical skills needed by employers in this rapidly developing field, also providing a grounding in the wider impacts on society and ethics.
Architecture-related content and demonstrations will explore the need for more solid data analysis in practice, including the use of AI systems to identify patterns and propose new solutions and strategies.
Ricardo Codinhoto, the faculty’s director for postgraduate taught courses, says: ‘Every company has a huge amount of structured data in CAD files from past projects which could be used to build a database for AI to analyse. Using this methodology we can detect patterns emerging in a practice's way of designing.’
Moving into concept design, generative design tools can identify patterns in strands of data and propose hundreds or even thousands of alternative dsign solutions. Codinhoto gives the example of using AI to carry out energy performance analysis of master plan designs to understand the combined impacts and compromises, ‘resulting in a better-informed decision making.’
In the field of project management, data validation in BIM is becoming increasingly important, with contracts set up to verify delivery at different project stages. An AI tool could interrogate the many thousands of points of data, says Codinhoto, and identify ‘what has been completed when, and to what extent, providing early warnings of missing information and triggering actions for those responsible to complete the work.’
Modules on the course with particular relevance to the built environment include ‘AI and sustainable development’, exploring the use of AI to tackle sustainability issues in areas such as energy, environmental pollution, circular economy, and decarbonisation.
Automation, manufacturing and design classes will look at the roles of big data, digital twins, the internet of things, and internet 5.0, as well as the use of generative design in architecture and civil engineering at the design stage.
The module Smart cities and the Internet of Things will focus on the collection of data through sensor networks and how the combination of generative design, urban planning and AI can help create sustainable and efficient solutions to complex problems.
Asked if architects should be concerned about their future roles given the sudden increase in AI’s capabilities, Codinhoto says no, not yet: ‘AI can considerably reduce the amount of time you spend on mundane tasks, like generating documentation, through automation. It is excellent at storing and processing data, but they are still pretty bad at coming up with new conceptual ideas.’ This, he says, requires human ‘common sense’ and an understanding of ‘context and where compromises needed to be made.’
Courses related to the development of AI systems already exist in the UK, but Bath’s is thought to be the first to train learners specifically in the integrated application of AI technologies in engineering and design.
Graduates can sign up either to a one-year full-time course or study over two years, and every student must complete a professional placement and an individual research project during the course.