In today’s rapidly evolving innovation landscape, AI promises transformative potential. What is this potential though? In a survey amongst our digital community, we dove deep into the reality. The report based on these results provides invaluable insights into how you can benefit from integrating AI into your innovation process and what are the main challenges to tackle. Let’s find out!
How AI is supercharging the innovation process
“AI shows tremendous potential to boost R&D”, you’ve probably heard this before. But what is this potential exactly? In our survey, we asked companies that are already using AI what is the impact is of integrating AI in different stages of the R&D process. In addition, we also compared this to the expected impact of companies who aren’t using AI yet.
In the figure below you can see the average scores on a 5-point scale of how impactful (from 1 – minimal to 5 – transformational impact) respondents considered these opportunities compared to the average usage rates. At first glance, we can see that the estimated impact of the opportunities is in line with the actual usage of them since the dots are relatively close to the baseline. Signaling that people are already investing in AI where they think it’ll have the most impact. Some, however, show some discrepancies.
Adoption of AI vs. estimated impact per development task ©Verhaert
The data reveals several key insights:
- Top activities like ‘test and develop code’ and ‘generate new innovation ideas’ are heavily utilized, leveraging AI’s ability to streamline repetitive tasks and spark creativity.
- Tasks such as ‘validate through surrogate simulations’ and ‘validate assumptions with virtual personas’ are valued but underused, indicating potential barriers to integration into workflows.
- Activities like ‘create pitch decks automatically’ and ‘validate assumptions with virtual personas’ are deemed less impactful, possibly due to the need for human creativity, leading to limited AI adoption.
- ‘Test and develop code’ and ‘validate through surrogate simulations’ stand out for their high impact potential, highlighting AI’s capacity to accelerate R&D by optimizing processes.
Overall, researchers see AI as beneficial for automating tasks and fostering innovation, yet skepticism persists regarding its effectiveness in creative endeavors.
Overcoming roadblocks in AI for R&D
No technology comes without its challenges. That’s why we also asked our respondents what their roadblocks to integrating AI into their R&D processes are. We also compared the input from companies that are already using AI to those who haven’t yet.
1. Challenges in AI adoption of AI users
First, let’s take a look at the results of the respondents who are already using AI. From the figure below, it’s clear that the biggest challenges are the lack of expertise and the lack of the right toolchains and methods.
Challenges in AI adoption rates for AI users per business size ©Verhaert
Looking more in-depth at the results, the challenges in AI adoption vary across different company types. Large corporations and start-ups cite a lack of expertise as the primary barrier, while SMEs struggle with limited access to toolchains and established methods for AI deployment.
The challenge of finding the appropriate expertise could indicate a need for more training and education on AI as well as a shortage of AI engineers and data scientists on the market. In the meantime, integrating existing AI talent, whether sourced internally or externally, into innovation departments could provide immediate solutions.
Corporate strategy gaps are particularly notable among larger firms. This suggests that some larger companies may not have a concrete plan for how AI will be integrated into their R&D processes.
Although less common, access to toolchains and methods appears to be a challenge, especially for SMEs. This could be due to resource constraints and less structured tools and processes. In addition, the absence of concrete projects hinders effective AI applications. Defined projects are crucial for identifying AI’s efficacy and establishing measurable KPIs.
2. Challenges in AI adoption of non-AI users
Next, let’s analyze the insights from the respondents who are not (yet) using AI. Looking at the total average, the main issue to overcome again seems to be finding the right expertise, followed by the lack of corporate strategy and toolchain and methods. These perceived or expected challenges could signal why they’re not yet using AI and be a growth opportunity for AI tools. Addressing the elephant in the room, there’s a clear difference in the answers of the start-up. Start-ups that aren’t using AI see the lack of a specific project and a corporate strategy as the most important causes. With the latter twice as important.
Challenges in AI adoption rates for non-AI users per business size ©Verhaert
3. Challenges in AI adoption AI users vs. no AI users
To conclude, let’s compare the results of AI users compared to non-users and evaluate the key similarities and differences.
Challenges in AI adoption AI users vs. no AI users ©Verhaert
For both groups, the lack of expertise (43% and 44% ) is the biggest challenge. This suggests that a lack of knowledge and skills is a major barrier to AI adoption in R&D in general.
Lack of corporate strategy (20% and 23%) is the second-biggest for both groups and shows a slightly bigger difference between AI users and non-users. Nevertheless, the difference between this challenge and the previous one is significant. This suggests that both those who are using AI and those who aren’t may struggle to get buy-in from senior management for AI initiatives.
Lack of toolchain and methods (20% and 21%) shows comparable results for both AI users and non-users, but rate it considerably less than challenges in expertise. This suggests that companies don’t struggle as much in finding the right tools and methods, but that even companies that are interested in using AI may encounter these challenges after they start the integration.
Lastly, the lack of concrete projects (16% and 12%) is the least common challenge and again shows a rather small difference between AI users and non-users. This suggests that most companies have concrete applications in mind when considering AI and that the biggest challenges lie in the actual application of it.
In conclusion
The data speaks for itself, the respondents agree that AI can have a clear positive impact on their R&D process, even those that are not yet using it. AI is poised to revolutionize the way we innovate. Nevertheless, to guarantee success, there’s a clear need for more training, education and for incorporating AI personnel into R&D departments, as well as corporate strategies for how AI will be integrated into their R&D.
