Industrial Sector

AI in product development: revolution or risk?

What opportunities does AI present in the product development and design process? Where do potential risks lie? Seven designers from Zühlke embarked on a self-experiment, yielding surprising results.  

6 minutes to read
With insights from...

The impact of generative Artificial Intelligence (Gen AI) on our daily lives became evident shortly after its emergence. The growth of this technology is expected to accelerate further, as illustrated in our recent blog post covering AI predictions for 2024.  

In product design and development, the use of generative AI is the subject of numerous discussions. The potential to speed up development processes and foster innovative solutions is weighed against concerns about quality and ethical responsibility. So, we decided to undertake a practical self-test to determine which prevails: the benefits or the risks.  

The results provide an insightful perspective on the opportunities and challenges of AI in product development, addressing questions like:

  • To what extent does artificial intelligence influence the product development and design process?
  • What potential does it have in terms of speed, efficiency, quality or creativity?
  • Where do potential risks lie?

Keep reading to discover our findings. 

Testing AI in product development

In our self-test, seven designers and engineers were given a practical design briefing to create a product concept with an associated value proposition. The briefing was based on a real-life customer scenario to maximise the test's relevance. Besides developing first-generation headphones according to customer design guidelines, the briefing included specific details on technology, target audience, personas, requirements, and time expenditure.

Six participants had access to various AI bots but were limited to one hour of processing time. The seventh participant opted out of using AI bots, acquiring significantly more time to pursue the traditional route.  

‘The product concepts generated by AI were innovative and convincing at first glance. However, textual and visual inconsistencies became apparent upon closer inspection.’

The initial results were astonishing. The AI-supported concepts, developed within just 60 minutes, appeared innovative and convincing at first glance: people and headphones were shown in a hyper-realistic visualisation, well-lit, and realistically textured, with no signs of stylistic or content uniformity.

AI-based vs. traditional product development: the devil’s in the details

However, upon closer inspection, textual and visual inconsistencies emerged. The most apparent were the incorrect size ratios between the head and headphones in Visualisation 1. 

Delving deeper, the representation of the eyes revealed a non-centred pupil, creating an odd appearance. In the detailed image at the bottom right, it is noticeable that the band is missing, rendering the representation nonsensical. 

on the left hand-side an AI-created woman with headphones and on the right hand-side illustration of the headphones with description of the functionalities and pieces. Visualisation 1: Initial visualisations are convincing at first glance but exhibit weaknesses in detail

This underscores once again: AI alone cannot replace the years of expertise and finesse of an experienced designer, but it can provide support. In our test, the customer-specific requirements were better met with the traditional approach, integrating the product into the customer's existing product world, translating brand values into the product, and making iterative improvements. 

Key learnings: AI product development considerations

AI in product development could be indispensable in the future. Yet, to ensure responsible practices, there are a few key considerations to keep in mind, as highlighted by our field test.

  • Reduced efforts and faster insights

    AI delivers breathtakingly fast responses to questions about users and the usage environment. However, accessible and reliable data is essential for meaningful and valuable results. Even then, manual verification of data is always necessary, as misinformation or misinterpretation can lead to incorrect and misleading outcomes.  

    Learn more about developing ethical applications with Responsible AI.

    Nonetheless, with the right data, AI can significantly accelerate research in product development and design, making it a valuable tool for increasing speed and efficiency. 

  • Trends prediction

    By scanning and analysing documented behavioral patterns, AI tools can identify future trends in user behaviour, serving as inspiration and optimising the product in terms of user-friendliness and user experience. This can significantly impact the product's market acceptance and success. 

  • Generation of innovative ideas and stimuli

    AI in product development enables the generation of innovative ideas or stimuli in large quantities and at record speed. As with generative AI in general, our test showed that the quality of ideas strongly depends on the input prompts. The results are most effective when initiated and controlled by a professional through targeted input.  

    The speed and volume of processed information and the resulting ideas are unimaginable, quickly leading to many innovative approaches. For us, this further demonstrates that AI is not the solution, but a tool that makes our work significantly more efficient. 

  • Faster visualisation of ideas

    The speed at which generative AI delivers convincing product visualisations is impressive – realistic product images are created in seconds. It might seem that neither product visualisations nor photographs will ever be necessary again. However, this impression is deceptive. 

    The test quickly revealed that iterative editing and detailed execution of ideas are exceedingly difficult. Our experience shows that initial renderings and sketches work well for creating moods and visions but beyond that, the quality drops. Final renderings or photographs created by designers cannot be replaced by AI-generated images of the same quality. The key lies in the optimal collaboration between humans and AI. 

    on the left hand-side a woman with white headphones, on the right hand-side a man with black headphones Visualisation 2: AI delivers first realistic visualisations in seconds
  • Quick and cost-effective process

    AI in product development has the potential to make processes faster and thus more cost-effective. However, further optimisation of the technology is needed to make the overall process more efficient. In certain subprocesses, a significant accelerating effect is already noticeable, but across the entirety of design and development this is not yet evident. 

    ‘The collaboration between humans and AI holds tremendous potential for the design and development process. The right mix of AI and human expertise is crucial for successful product innovations in the future.’ 

AI in product development: the right mix is key

Our field test further underscores that artificial intelligence will not replace the traditional product development process. Instead, AI is a support tool to enhance it. 

Ultimately, it's the combination of AI, quality data, expert operation, competent evaluation, and further processing of outputs that enables designers and product developers to optimise processes or outsource entire tasks. 

For example, simple, repetitive, and time-consuming tasks can be automated and completed faster with the help of AI. In the area of data-based research, AI is already significantly superior to traditional methods. 

The collaboration between humans and AI offers enormous potential to scale and accelerate design and development processes, save resources and costs, foster creativity, and enhance the customer focus of products. The right mix of AI and human expertise is crucial for successful product development in the future. 

Zühlke has extensive experience in industrial design and product development. Together with our clients, we leverage new technologies and trends to create innovative products and solutions. So, if you’re thinking about creating a new product or need help with an existing project – don’t hesitate to reach out to our team. We look forward to hearing from you. 

Contact person for Switzerland

Philipp Morf

Head AI & Data Practice

Dr. Philipp Morf holds a doctorate in engineering from the Swiss Federal Institute of Technology (ETH) and holds the position head of the Artificial Intelligence (AI) and Machine Learning (ML) Solutions division at Zühlke since 2015. As Director of the AI Solutions Centre, he designs effective AI/ML applications and is a sought-after speaker on AI topics in the area of applications and application trends. With his many years of experience as a consultant in innovation management, he bridges the gap between business, technology and the people who use AI.

Thank you for your message.