Welcome to the Age of the Engineer-Data Scientist
The growing enthusiasm for a new hybrid role raises significant questions. We answer them here.
- By Brett Chouinard
- March 3, 2022
The typical product development/simulation engineering team now enjoys access to a wealth of data that can and should be informing their product design and manufacturing processes. However, finding critical insight within these vast reservoirs of information is another matter. New skill sets are urgently needed. Specifically, engineers must be able to harness artificial intelligence (AI) and machine learning (ML) to support and accelerate better decision-making.
This fundamental shift is illustrated by the emergence of a new hybrid role -- the engineer-data scientist. What’s more, the success of these multi-skilled pioneers will be crucial to the future of the enterprises that are recruiting and training them. Ultimately, engineer-data scientists must shoulder the task of turning the undisputed potential of AI and ML into faster time to market, and they must design more efficient products that perform better for customers and end users.
It’s a big ask, and the growing enthusiasm for the new role raises significant questions. Are engineers really the best people to pick up the data science baton? If they are, what skills do they need? On a more practical level, how can they acquire the mindset and capabilities of data scientists? What are the implications for organizations?
In engineering and beyond, the data science revolution is gaining traction. A recent PwC survey reports that 86 percent of respondents describe AI as a mainstream technology within their organization. However, in many respects, we have only scratched the surface. A Capgemini report reveals that by deploying AI at scale, automotive OEMs could increase profitability by 16 percent. There is frustration, too. The aforementioned PwC survey also notes that 76 percent of organizations are barely breaking even on AI.
In the search for better return on investment, the creation of the engineer-data scientist is a significant landmark. It reflects growing recognition that solutions should be driven by domain expertise. In other words, the people with granular understanding of the metadata and engineering challenges are the best people to apply the tools that will uncover insight and can thus navigate the best route forward.
Are engineers a good fit for the role? There are convincing arguments in their favor. To start with, although the impact of AI and ML will be revolutionary, it also represents an evolution from what has come before. There are clear parallels with the principles of established engineering techniques such as experiment design, as well as modern simulation and optimization tools. Across every discipline and sector, engineers are comfortable working with simulation, analytical modeling, and statistics.
A Small Step
Of course, the scale and speed at which AI and ML work (and their unprecedented ability to embed continual learning) are revolutionary. At the same time, given their existing capabilities, most engineers will find that embracing data science is a small step rather than a giant leap. By nature, engineers are curious and thrive on solving problems. Moreover, they do not work in the world of pure science. Their focus is on delivering commercially viable solutions. Ultimately, engineers are motivated by a practical desire to build something better. Instinctively, they will be drawn to tools that can help achieve this goal.
Engineers adopting data science are greatly helped by the latest low-code and no-code AI and ML tools. Democratization is hard at work, and the workflows are increasingly familiar and intuitive. However, prospective engineer-data scientists still need to develop new skills. The encouraging news from universities is that data science is an increasingly popular option in engineering courses. Given the urgency of the requirement, we will also see plenty of on-the-job training for more experienced engineers.
Where do I start? is probably the most common question we hear from aspiring engineer-data scientists. The short answer is with algorithms. AI and ML are essentially about matching and applying the right algorithm to the right problem. It is extremely unlikely that an engineer-data scientist will have to take on the job of actually writing these algorithms.
Beyond this, engineers should be encouraged to get involved in projects where they will use AI and ML. From here, we can be confident that their inclination toward hands-on learning is an ideal springboard for a new wave of data science advocates.
This largely organic career path means that engineer-data scientists are always likely to prove an easy fit with their colleagues and organizations in general. Specialist data scientists will remain an important piece of the puzzle, scaling solutions developed by domain experts and building the necessary infrastructures. The difference made by the engineer-data scientist will be seen in design and manufacturing outcomes, not corporate restructures.
If further evidence were needed of the suitability of engineers to take on these new responsibilities, it can be found in a data science sector that is recruiting engineers to fill its own skills gap. Hopefully, the engineer-data scientist role will mitigate the risk of a brain drain. For anyone interested in smarter, more sustainable products, we need our engineers to keep on engineering. We also need them to turn their talents and attention to getting the very best from what data science offers.
Brett Chouinard is the chief product and strategy officer at Altair, where he is responsible for the strategy and vision of Altair products, which includes facilitating the development, sales, and delivery of Altair’s solutions. You can reach the author on LinkedIn.