I recently had the opportunity to sit down with Shirish More, the Product Manager for Siemens NX, to discuss how NX is implementing artificial intelligence (AI) to enhance and enrich the mechanical design process. In the first part of this series, available to listen to here, we talked about how AI leverages the wealth of engineering knowledge inherent in completed designs to support the sharing and transfer of knowledge within design groups and between people. I would go on to learn from Shirish that he and his team were focused on three key areas: knowledge capture and transfer between long time engineers and expert users. Uncovering patterns across a wide array of sources, as well as sharing and recommending workflows based on accumulated knowledge.
At the beginning of our talk, Shirish offered a great definition of AI: “Artificial intelligence – especially when it comes to mechanical engineering software – is our ability to recognize design patterns and solutions.” This ended up setting the tone for much of what we would talk about over the next 15 minutes. In mechanical engineering and design one of the greatest strengths an individual or company can have is the ability to draw on decades of past designs when facing a new problem. This is a quite literal application of the old saying “don’t reinvent the wheel” since for many engineering problems a well-developed solution already exists which can be easily adapted to fit whatever constraints a designer might be working under.
Traditionally, engineering knowledge is contained within textbooks, guidelines, and completed designs which places the burden of learning solely on individual engineers. While every engineer has a baseline of skills and relevant knowledge to draw on, no single person could be expected to have complete understanding of decades of completed projects. This is exactly where Shirish and his team are using AI can help bridge the gap. By treating engineering knowledge as training data, it is possible to encapsulate all the practices and solutions inherent in past designs within a system that intelligently offers suggestions, presenting the information in an accessible format to engineers of all experience levels.
Design data and documentation are not the only sources of potential training data for AI in mechanical engineering software however. Long-time engineers and expert users also represent a valuable source of data which an integrated AI can learn from in real time as these veteran users drive the software. This technique is another area where Shirish and his team are working to implement AI method for retaining valuable expert knowledge that would either be difficult to record or lost entirely when an engineering moves on from a position or retires. Shirish explained that by capturing such a wide array of knowledge from disparate sources and consolidating it into a single system, it becomes easier for engineering teams to work as true unified groups where all members have equal access to information, regardless of experience level. Beyond sharing solutions, an AI can highlight or uncover patterns present in data across a wide array of sources, providing deeper insight or even improving design practices.
Beyond the broad recognition of patterns and solutions, Shirish went on to explain how an AI system like this could also share workflows between users. In his example, an experienced user is given a task which they perform quickly and efficiently since they are highly familiar with the tools. A few months later, a less experienced user has to perform the same task. In this example, the newer user would likely need to consult documentation or receive guidance from the more experienced user. While neither of these options is inherently bad, they do take time. Now, however, a third option is available. Thanks to the AI learning the efficient workflow of the experienced user when the task was first completed, the new user can be provided a sort of “real-time training”, learning to complete the task as they complete it under the guidance of the AI system. In this case, the AI assistant takes the vast knowledge it has been trained with and distills it down into pointed suggestions and recommended workflows. Implementing this smoothly and seamlessly has been a challenge for Shirish and his team to overcome.
During my talk with Shirish, we covered many exciting topics in AI, especially in how it can help expand the design space and fundamentally change the way we interact with mechanical engineering and design software. If you are interested in not just the way AI is preserve and share engineering knowledge, but also how AI is pushing the boundary of smart human-computer interactions to become a true design assistant, stay tuned for part 2 of this podcast series. To listen to part 1 of this podcast series, click here. To read the full transcript, click here.
Siemens Digital Industries Software is driving transformation to enable a digital enterprise where engineering, manufacturing and electronics design meet tomorrow. Xcelerator, the comprehensive and integrated portfolio of software and services from Siemens Digital Industries Software, helps companies of all sizes create and leverage a comprehensive digital twin that provides organizations with new insights, opportunities and levels of automation to drive innovation.
Siemens Digital Industries Software – Where today meets tomorrow.