Along with Artificial Intelligence (AI), ‘smart product’ seems to be one of the buzzwords of the decade. But developing a smart product is a complex process, requiring close collaboration across the whole value chain. In this article, three imec experts share their expertise in smart innovation. The main takeaway? Developing a good product takes time, don’t rush the initial innovation stages!
To be smart or not to be smart?
Smart products seem to be everywhere, but what exactly is it that makes them deserve this ‘smart’ label?
Kris Van de Voorde, innovation manager at imec, explains: “There are different degrees of ‘smart’. An advanced smart product could be a connected device that collects data, which is then used as input for AI-based operation. A driverless car, for instance, is an example of such an ultimate smart product, as it is connected to the cloud and equipped with various sensors and radars, which deliver input for AI-based decisions. But a product can also be called ‘smart’ without an external connection or without the use of AI. And although we tend to associate the term ‘smart’ mainly with products, processes can also be made smart. In a factory, for instance, it might be worthwhile to add intelligence to your production process, so machines can predict when one of their components is likely to break down, allowing you to anticipate this and replace the component without having to halt the entire production process.”
In the last few years, the consumer market has become flooded with gadgets claiming to be smart – ranging from smart security cameras and smart speakers to smart toasters and even smart baby changing pads. But it’s not because it can be smart, that it should…
Kris Van de Voorde: “Developing a smart product is much more complex than developing a regular device. It takes collaboration across the value chain and an enormous amount of time, money and expertise. So, the first question you have to ask yourself is always: is anyone actually waiting for this product? Will the effort be worth it?
Tanguy Coenen, application prototyping team manager at imec:
A staggering 90% of digital and hardware innovations fail and the most commonly stated reason for this is simply that there is no market need for it.
"At imec, we have a whole team specialized in user research. By involving end-users right from the start, they can accurately map their needs and help companies estimate the market value of their idea. In the next stage, my team develops application prototypes that we present to users. We let them experiment and then – based on their feedback – develop a second version. We usually do several reiterations before coming up with a concept that is a good problem solution fit and matches users’ expectations.”
When hardware meets software: a tricky, but promising union
Products are becoming increasingly complex, containing more mechanical and electronic components, but also more software codes. Due to this complexity, they often require a multi-domain approach, integrating mechanical, electronical and other hardware and software. Not an easy feat as the electronics supply chain tends to be quite fragmented and separated from the mechanics and software supply chains, with different companies and teams mainly focused on their own field of expertise.
Kris Van de Voorde: “When imec merged with iMinds in 2016, we experienced first-hand that hardware and software experts operate in different worlds, using different terminology and methods. I remember long discussions just to come to a shared definition of the term ‘proof of concept’, which apparently has different connotations in both fields. Now, of course, having both these hardware and software experts on our team has become a big advantage. We’ve gained experience about how to collaborate across domains and have diverse expertise in-house, allowing us to help companies with more than one aspect of their smart innovation puzzle.
To build successful products, experts have to venture out of their comfort zone and really collaborate, learning from each other’s methods.
With their smart steamer and related software application, imec.istart company Mealhero illustrates perfectly how smart products require design, hardware and software to be perfectly aligned.
Look before you leap: the importance of the “shift-left” in product validation
When developing a new product or system, it is tempting to focus mainly on functionality: does it do what it is supposed to do? But if a product cannot be properly produced, is not reliable, or is simply too expensive, it is unlikely to be successful. It is, therefore, important to test your innovation beyond only functionality and – ideally – to do so as early on as possible.
Geert Willems (imec - Center for Electronics Design and Manufacturing): “As products become more complex, it becomes increasingly important to validate design options as early as possible to avoid costly and time-consuming re-designs in a later stage. This is called the ‘shift left’: you move validation to the left in your product development timeline. So, instead of building a physical prototype when the design is completed to test all aspects of it, you try to validate as many aspects as possible before building an actual prototype. It is also critical to integrate product requirements such as manufacturability and reliability into the design phase. It should not be an add-on in the industrialization phase when design changes have a major cost and time-to-market impact.”
“In the IC world, this is already common practice due to the huge costs (million dollar mask set) of a design change. But in the wider electronics domain, the focus is still very much on building and testing physical prototypes. If the prototyping tests are not satisfactory, they need to redesign and build a new prototype . This can delay the developing process by weeks up to months. Many things that are now tested in this way – for instance, related to the product’s manufacturability, robustness and reliability – can be validated in advance by Failure Mode and Effect Analysis (FMEA) and virtual prototyping (simulations).
Detecting and fixing problems at an earlier stage is much more time- and cost-efficient. As a rule of thumb, the cost for changing a product design increases by a factor of ten at each crossing of a development stage-gate.
Some examples of companies we’ve supported
Morrow develops autofocal glasses, that switch focal length electronically. They use the imec cleanroom for their low-volume production and collaborate with us to tackle technological problems and to work on the certification of their electronics.
Mealhero launched a smart steamer with connected app to provide easy, but healthy meals. As an imec.istart company, Mealhero benefited from both business and technological support provided by imec. E.g. our researchers helped them make sure their IoT hardware platform and software cloud platform are perfectly aligned.
Cochlear develops hearing implants that bypass the outer and mid ear in order to directly stimulate hearing cells. They collaborate with imec to further miniaturize their technology, so it matches human physiology better.
Want to know more?
Imec’s Innovation Services and Solutions brings together the different types of support and expertise we can offer throughout the innovation lifecycle. For more information, check out our website.
Imec’s Center for Electronics Design & Manufacturing supports the development of high-quality, reliable and cost-effective electronic modules in the industry. For more information, take a look at their website here.