Complex systems—which often include computing capabilities, mechanical systems, and electronics—are particularly challenging to design. Artificial intelligence can be used to assist engineers throughout the design process, and less-critical design tasks can even be automated.
Artificial intelligence (AI) has a growing role to play in business processes, especially product design. When it comes to complex systems, specialists from a wide range of fields must work together, sharing information and providing input throughout the design and simulation processes. The researchers assigned to our INCA program are working on two key practical applications for AI techniques in the fields of software and systems engineering.
AI techniques can be used to automate a certain number of activities and tasks at different stages of the design process. In the earliest stages of project development, for example, AI can be used to help formalize a problem and, potentially, reuse solutions from previous projects and even from completely different industries. Learning algorithms—machine learning and deep learning—and expert systems can be used to build design automations companies can use to make their processes more efficient.
This approach is particularly helpful in software engineering, where open-source code repositories and other resources are readily available, serving as huge knowledge bases. Researchers can also extract knowledge from resources such as Wikipedia, scientific article repositories, and industry-specific ontologies to bring these learning algorithms to new engineering use cases.
In addition to these automation tools, our INCA program is developing specialist chatbots for systems engineering. These virtual assistants support engineers throughout the product development process. Once the chatbot is launched, it stays open on the engineer’s computer, analyzing work in real time, making suggestions, and highlighting potential issues. The engineer can also ask the chatbot for advice or suggestions based on similar projects developed within the company or even in other industries.
Natural language processing (NLP) facilitates chatbot–user interactions. Users make requests—like describing the product they wish to design—in natural language; the chatbot analyzes the text, and then generates an architecture model corresponding to the user’s requirements. This model can then be used as a basis for further development by the engineering team.
