Applying AI in engineering…
Much of the excitement around Gen AI has focused on the possibility of early results – on its ability, for instance, to write an executive summary of large amounts of information.
However, AI’s capabilities shouldn’t be considered as a series of isolated decisions or small-scale automation projects to address local process inefficiencies, but should rather be seen as an opportunity to step back, holistically assess the value that the business process should create, and seek a global improvement, by redefining activities from the ground up.
Engineers have this kind of pragmatism. It’s a question of attitude – of the ability to think strategically as well as tactically (see ‘Capgemini Engineering’s mindset’ at the end of this section). That’s why we recommend that, in addition to being happy to take an early result, they should also be looking further ahead. In our view, it makes sense for engineering organizations to look for ways to create value over different time horizons for different levels of transformation.
… in the short term…
Let’s look first at generating short-term value on the bottom line. This is where the ability of Gen AI to digest and synthesize complex information comes in. Combined with Retrieval-Augmented Generation (RAG), it enables engineers to use simple natural language routines to search and query knowledge bases or technical documentation. In manufacturing, this same Gen AI ability can accelerate classification and analysis of quality incidents. It can also accelerate generation of quality documentation in domains with heavy compliance requirements (such as life sciences) and automate the documentation of non-conformities. You can find other use cases here.
… in the mid term…
In the mid term, artificial intelligence can generate broader, deeper value on the bottom line. This is the domain of Hybrid AI, which brings Gen AI together with symbolic AI, reasoning engines, logic routines and more, and may take benefits beyond the point-tasks we have just seen, to broader classes of engineering – focusing, for instance, on augmenting the end-to-end scientific discovery process, rather than on summarizing research papers. It’s also at this point that we can see intelligence becoming an integral part of products and services across organizations – from the cloud to the data center, from the office or factory floor to the edge.
… and in the longer term
In the longer term, Hybrid AI starts to change the nature of engineering in ways we may not yet be able to imagine, and in doing so, begins to deliver significant value to the top line. For example, it may absorb knowledge of industry tooling, processes, and norms from across the engineering organization, and use that knowledge to suggest entirely new products and processes, and new routes to achieving them cost-effectively and sustainably. It’s at this point that Hybrid AI moves beyond ‘doing the same thing, but better’, and steps into exciting uncharted territory.
Capgemini Engineering’s mindset reflects the lessons learned from its extensive experience of running engineering projects. It’s a mindset that aims to be time-responsive and context-sensitive, ensuring it remains relevant in an ever-changing technological landscape – and particularly in relation to the role of Hybrid AI in Augmented Engineering.
The mindset addresses several key areas:
1. Artificial intelligence is transformational – a piecemeal implementation of AI, including Hybrid AI, will not succeed. We should seek comprehensive adoption of AI across the organization, focusing on transformational applications rather than isolated use cases
2. Treat sustainability as a systems issue – prioritize according to reliable data; integrate sustainability into every aspect of engineering practices to ensure long-term viability and
environmental responsibility. Also, rational and responsible decisions about environmental impact must be made at the system level. For example, the impact of training and using a Gen AI model may be a tiny fraction of the benefits gained from reducing weight or energy consumption across the life of an aircraft or a truck
3. Agility through model-based engineering – exploit collaborative data models and mature tools to accelerate delivery and acceptance across disciplines, enhancing responsiveness and adaptability in project execution
4. Prioritize the role of the human in engineering – AI will transform products, but both its ultimate users and creators are human. Creativity and problem solving in development should be incentivized, while also placing human-centered design and social acceptability at the forefront of product development
5. Engage with key ecosystems – the transformative benefits of AI will cross organizational boundaries, contributing to the standards and tools that enable network effects
