Engineering leaders are always looking for that next competitive edge to make their product more efficient, higher performance, or lower cost. For the past couple of years AI has been a key point of interest in terms of the possibilities it could bring to the engineering field. With NVIDIA temporarily peaking as the world’s most valuable company “fuelled by an artificial intelligence craze that has triggered wild demand for the company's chips and graphics processing units” (Forbes, June 2024), AI is more prevalent than ever.
As a leader in engineering, you've probably been following AI developments, discussing them with peers, and considering their application in your workflows. But if you look back on the past six months, have you achieved your AI adoption goals?
Companies have experienced varied outcomes with AI. Some, who began AI projects years ago, are now enjoying cost reductions and faster deliverables. Others have gained valuable insights from larger AI implementations. Even those just starting to focus on machine learning are upskilling their teams and positioning themselves for future successes.
On the other hand, some companies remain cautious, preferring to observe from the sidelines and waiting for guaranteed benefits before launching their own initiatives, often years behind their competitors
So, what approach works best?
We've collaborated with engineering teams across various sectors for the past six years. We have guided projects of all sizes and risk levels. This blog highlights why you must seriously consider AI implementation now to avoid falling behind your competitors.
Delaying AI initiatives costs more than just time or money — it risks your market relevance. Early adopters gain significant, often invisible benefits, like streamlined processes and competitive advantages, that put them miles ahead. These benefits do not make headlines but rather allow the company internally to position themselves better for the coming years. From the outside, the company appears to have quickly integrated a new technology with their business. The reality is that the efforts start months and years in advance. Building up your data management strategy
Many firms recognise AI's importance but fail in execution. They commonly make the mistake of launching ambitious AI projects without a foundational strategy. For successful AI implementation, you may need to adjust fundamental engineering processes, such as data storage and management. These changes can introduce overhead and slow progress.
As an example, we have clients with large production databases of test results. They have terabytes of test data available that could be used for a large AI project, but that is not where they start.
Instead, they begin with a smaller use case. Looking at past experimental data that is no longer being used or picking a very specific test. As an example, this could mean looking at data from a few years ago that is no longer current. Or if testing in a thermal chamber – using the data from a small subset of the data.
You can learn a lot of lessons from smaller-scale pilots that will be just as relevant for a large-scale AI project with very little risk involved. This can relate to how you structure your data, find gaps in your testing processes that you may need to fill, and reevaluate the validity of certain experiments, etc.
Risk: If a competitor has already learned key lessons from AI from a small project, they will be able to successfully scale very quickly to larger implementations.
Actionable Tip: Start with a smaller AI Project and build up good habits in terms of data management such as file storage, file naming conventions, data archiving, access controls (and so many more things that you can only learn from implementing a project yourself).
Even if you're not implementing AI today, prepare your teams for its future arrival. Engineers need to bridge typical engineering processes and machine learning expertise. Training your engineers in machine learning now can pay dividends later.
Not only will you have more skilled team members, but you will build practical experience for the job: A combination of understanding of current processes and limitations as well as machine learning expertise.
Some companies may hire data science experts into their team, but data scientists and engineering team members should not be siloed in their knowledge. Training your engineers for AI use cases makes it easier for them to communicate, understand limitations, and identify better use cases.
Risk: If a competitor who has upskilled their engineering team with machine learning, they are more likely to understand the implications early on and be able to anticipate opportunities for improvements, risks, issues, etc.
Actionable Tip: Invest in training your engineers in machine learning, even if your AI journey hasn't started. This preparation will position your team for future success.
Implementing change at an organizational level presents significant challenges. The transition to AI can seem particularly daunting, especially from a risk perspective. Mainstream media often highlights various AI use cases—some more successful than others—creating preconceived notions that might hinder AI adoption. Starting with a small project can help build company culture, increase risk tolerance, and foster a more open and understanding attitude towards the possibilities of machine learning
Furthermore, once you have a successful small scale AI project, it can be used as reference/guidance for a future project to build upon which can serve as a business case/internal case study.
That way, you can translate what a larger AI project could mean based on previous direct experience from your specific environment.
Risk: You may face hurdles and delays in starting a large AI project if your company has not been exposed to similar projects in the past. Comparing this to a competitor that may have previously started a smaller AI project, they will now have a culture where they are more accepting and open to a large-scale project and can get to market faster.
Actionable Tip: Going through the change management process for a small project can help you identify blockers, potential issues, and prepare yourself in advance of when you want to start on a larger AI project.
Future benefits of AI Adoption
Imagine your firm in two years, with AI seamlessly integrated into key areas. Testing times slashed, costs reduced, and your R&D team is innovating at an unprecedented pace. The rest of the world is adopting AI, some at a faster pace than others.
Not only is this possible, but a lot of large engineering firms are also on track to achieve these goals.
The cost of inaction
Recently, we had the pleasure of hearing from many engineers at several tradeshows, including Battery Cells and Systems, Automotive Testing Expo, and The Battery Show Europe. They shared insights on how they started various AI projects, the different approaches they took, what has been successful, and the key obstacles they faced. These discussions highlight the popularity of machine learning in the engineering domain.
Such conversations will only become more frequent in the coming months and years. Companies that began their AI journeys years ago are now reaping the benefits. Their engineers understand the power of machine learning and its business applications. They started with small projects, gradually expanding and balancing experimentation with acceptable risk. Now, they have gained experience, upskilled their teams, and built a culture that embraces these modern technologies.
In our July 2024 webinar on the state of AI in EV battery validation, one of our panellists explained how other companies already leverage AI and pointed out that many firms are farther behind than they think.
Conclusion
Starting your AI journey now prepares your company for a future where AI is
integral to success.
Planning your AI implementation in the coming months? Check out the link below to learn more. Our examples of case studies where other engineering firms applied AI will showcase some possibilities that you may want to adapt yourself.