b'28 Partner News Summer 2023CorporateLarge LanguageAI and LLMs can be used to automate many of the more time-consuming aspects of engineering design. For example, an LLM could be trained to automatically generate Models: RevolutionisingCAD models based on natural language inputs, reducing the time required to create engineering deliverables. However, its Engineering Designcapabilities and potential go much further than this.LLMs could be taught to understand the meanings of codes, clauses, formulas, and standards, which allows engineers to develop an algorithm that can automatically put together calculations By James Crouch, Technology Director and Brian Palmer,based on design requirements. LLMs can then produce easy to understand technical Digital Delivery Engineer at Burns & McDonnell descriptions and suggestions, just as a human would. In essence, they will act as design assistants to engineers, able to make informed decisions and apply codified standards.It is important for the sector to consider how these language processing capabilities can be leveraged, so that LLMs are adding value to projects by identifying potential risks, optimising designs and providing cost-effective solutions. One of the most significant benefits of LLMs is their ability to generate multiple potential solutions quickly and efficiently. Engineers can use these as a starting point, exploring designs further or making refinements.The landscape of engineering designLLMs can also help engineers incorporate has seen significant change over the lastinnovative solutions that are not decade, but recent advances in artificialimmediately obvious; the vast amounts of intelligence (AI) and machine learningdata they are trained on mean that they shows that there is far more potential forcan identify patterns that are not, at first, technology to enhance work in the sector.apparent to humans. This means they have Recent advances in large language modelsthe potential to identify materials and (LLMs) such as ChatGPT are evidence thatconfigurations that go against the grain there are many more exciting applications toof conventional designs, but that offer introduce new and innovative approaches tosuperior performance.engineering design.LLMs will enhance the role of engineers LLMs are AI models designed to understandby giving them more space for creative and generate natural, human-like textthinking and problem solving, and reducing based on large amounts of input data,the time they spend on less cerebral tasks. and many people will have seen examplesEngineers should be conscious of the asking the programme to write an essay orneed for vast quantities of high-quality a sonnet. However, their potential extendsdata to make LLMs worthwhile, as well as far beyond simply generating text, asthe need for effective human oversight they could be used to analyse, interpret,and expertise. Nonetheless, it has the and even generate complex engineeringpotential to revolutionise the engineering documentation and instructions. The modelsdesign process and allow clients to access have made significant progress very quickly,more efficient and optimised designs and so engineers should start thinking aboutsolutions. Harnessing the power of LLMs how best to integrate LLMs into their designwill allow engineers to unleash a new era of process. innovation.'