Interview: Visteon – How We Create HMIs for Connected Cars

In our interview series, we talk with people and companies that are at the forefront of creating the digital connected cars of tomorrow.

In our first interview, Laurent Vioujas from Visteon shares his thoughts on the digitalization of today's connected car. Laurent also talks about how they've made the transition to digital dashboards and how digitalization has impacted their development process.


1. Introduce yourself and what you do

Laurent Vioujas of Visteon on creating HMIs for connected cars

My name is Laurent Vioujas. I am a UI Software Lead engineer here at Visteon Corporation, which is one of the largest Tier 1 automotive suppliers. I live in Detroit and have been working here for the last seven years. I have had the opportunity to work both on production programs and Visteon’s innovation products. Our teams use Rightware’s Kanzi UI Solution for creating automotive Human-Machine Interfaces (e.g. clusters, infotainment systems and HUDs) for our clients. Visteon is also participating in Rightware's Kanzi Partner Program.

My team is currently working on high-end, fully reconfigurable instrument clusters that will go to market in the next few years. Our team is global with team members in North America, Europe, and Asia.

2. How has the transition from traditional instrument clusters to digital clusters changed the development process?

The transition from traditional instrument clusters to digital clusters has been progressive for me. We’ve gone from working with segmented displays (“traditional” physical gauges with an LCD in the middle), to LCD cluster displays - 4.2”, 10.1” with ever-growing resolutions.

With segmented displays there tends to be less flexibility. The design has to be discussed and agreed upon upfront, and each change tends to require quite a bit of cost and work, which means that the design is generally frozen pretty early on in the development.

The move to digital instrument clusters was a learning process – both for the customer and for us! We found that we needed to educate customers on GPU limitations, 2D/3D capability, gradient rendering, font rendering, and the entire change control process that’s needed when developing digital interfaces. With a digital cluster we have a large surface to work with, which allows for more flexibility and more innovative design. Rightware Kanzi makes it possible for us to implement design changes rapidly.

It’s a bit of a new world for everybody, including the customer and our HMI teams. However, at Visteon we have worked hard to educate everyone involved and make sure our stakeholders understand that each GPU is different and has different capabilities. Most of the time you tend to get what you pay for, but when you use a proper graphics solution, there are great graphics effects that can be achieved even with cheap GPUs!

Lightscape 1

 Image: Visteon LightScape™ reconfigurable instrument cluster, created using Rightware's Kanzi UI Solution

3. In your opinion, what is the most challenging part in the process of developing an automotive HMI?

An automotive HMI isn’t something that gets updated on a monthly basis. This might change in the near future, but for now when you buy a car at the dealership the cluster comes with an HMI which you are going to keep until you change car. When you compare it to your smartphones, computers or even now smartwatches which can be easily skinned or customized this is very different. There are so many people involved in the development of an automotive HMI, and each person has their own preferences, so it can sometimes be challenging to come to a consensus on design. Some people prefer dark colors, others like larger fonts, 3D models, etc. The HMI you see in your car has gone through numerous iterations during its development and many people have voiced their opinions. As a consumer, what you see is the final product.

To answer your question, the most challenging part in the process of developing an automotive HMI tends to be the establishment of the final requirements; the final decision where the graphics and the user experience has been agreed upon by the entire approval chain at the OEM.

Rapidly prototyping our design on real hardware tends to be one of our most effective tools for eliciting requirements. Showing designs on a projector screen doesn't give an accurate sense of the product; the large image hides nuances in the design, so reviewers tend not to give many comments. Instead showing the design on a 12” digital cluster behind a wheel, completely changes their mindset! That’s when they start noticing even the small things and the feedback starts coming in. Rightware Kanzi has been an essential tool for us in establishing a standard process for rapid prototyping.

4. We see the Engineer and Designer roles becoming closer, even merging in some cases. How do you see these roles evolving?

In automotive HMI development, I think that experts are still needed on both ends. We need expert engineers in software programming and experts in design, but I also see a third role: a "Designer-Engineer" capable of performing most programming tasks as well as graphics rendering. This person can perform most day-to-day activities with modern HMI tools like Kanzi, but needs to rely on the experts in some cases where something really complex or outside of the ordinary needs to be done. Each person is different and while some people can excel in a very specific domain (C++ programming, shader programming, 3D design, etc), others can be pretty good in all of these domains! All of these roles bring value to a larger complex project like a custom automotive UI. We need the experts early on in the development of the project to set the architecture, the structure, and the rules, once all of this has been established they are not needed as much and can move on to their new projects while still providing ad hoc support during development. Of course, you need support for the technology you are using as well. With Rightware Kanzi, we leverage the local Rightware team here in Detroit which has been extremely helpful for us.

5. How does the end user benefit from a well-made HMI?

A well-made HMI should be intuitive and help the driver stay safe - not distracted or confused. For now autonomous driving is still a few years away, and the main job of the driver while on the road is to drive safely. The end user should be able to easily figure out the logic between the various menus he has to access on his steering wheel. Consistency is the key. Unfortunately, it can be easy for engineers to create exceptions in the UX which end up being confusing. For us here at Visteon, user experience is extremely important which is why we often work very closely with the customer in making sure that the UX and the actual implementation stay aligned.

6. What do you think is a good skill set for becoming an automotive HMI designer or engineer?

This may sound like a cliché, but it is really, really important to be able to think outside the box and to think about the “abnormal” use cases. Unfortunately, these are the use cases that tend to take 99% of our time. We frequently ask ourselves during our meetings, “What should the cluster display do if A, B, and C all happen at the same time?” Even though these scenarios represent an extremely small percentage of the use cases and the end user might never see it in the years he will own his vehicle, we need to be able to handle those properly to ensure safety. Being able to spot these cases as early as possible in the design process is a key skill for both engineers and designers.

7. How do you see the role of the Tier 1 evolving in the future?

Tier 1s currently have the role of delivering an embedded electronic product combining hardware and software according to the OEM’s specifications. The Tier 1 is responsible for selecting the right hardware/software eco-system to address these requirements in terms of features, quality, time-to-market and cost.

While some OEMs have started to define the HMI - from graphic design to platform software delivery – for the Tier 1 to integrate, others prefer a complete solution designed and supplied by the Tier 1.

Tier 1s are well positioned to observe the market developments and is a good source for new ideas providing product and market differentiation to the OEMs. Technology innovation and the creation of a unique user experience require a tool for the Tier 1 and OEM to align the strategy and collaborate across the increasingly capable (but more complex) graphics eco-system. Using the most optimized tool for graphic design and platform execution is therefore essential to enable the creativity in the collaboration process, while reducing requirement capture and development cycles.

8. How do you see automotive HMIs such as clusters, infotainment systems and HUDs changing in the next 5 years?

With prices of powerful embedded 2D/3D GPU going down, some OEMs will probably go down the path of having one unique HMI for their entire vehicle line with maybe a few exceptions for their flagships or performance vehicles. Others may go to the opposite extreme with “skinnable” UIs for each model. Again, we need to remember that safety is the key priority and we cannot be distracted by pretty, flashy skins. I am hoping OEMs will use their digital instrument clusters, IVIs and HUDs more as selling features for their vehicles in the future because, at the end of the day, you spend most of your time watching the road when you drive, but the instrument cluster is next.


Thank you for the interview, Laurent.

The Future of Automotive UX - Download presentation deck