How platform thinking drives innovation at Visteon

Raman Mehta, CIO, discusses the shift from traditional manufacturing to software engineering at one of the world’s largest cockpit electronics companies.

automation gears lightbulb machine learning ai innovation by bananajazz getty
bananajazz / Getty Images

When software and data become an increasingly significant part of your manufacturing company’s products, how does your product development organization need to change? How does the IT organization, trained to enable mechanical and electrical engineers, now enable a 5,000-person software engineering organization?

These are the questions that faced Visteon, the automotive parts supplier, when it moved headlong into the cockpit electronics market several years ago. Raman Mehta, who became CIO in 2017, bases much of the transformation on platform thinking, which allows IT and engineering to focus on innovation, not just the basics.

Martha Heller: How is Visteon transforming as a company?

Raman Mehta: A few years ago, we made a conscious decision to shift our focus from automotive climate control and interiors to cockpit electronics.

Raman Mehta, CIO, Visteon Raman Mehta, CIO, Visteon

To deliver a digital driving experience, we are producing infotainment and driver information products that have multiple domains combined into one software stack. We need to consolidate driver information, driver monitoring and infotainment domains into the vehicle — or what I call a computer on wheels — to deliver a consistent user experience.

This means that the products have become much more configurable and complex and we needed to shift from mechanical and electrical engineering to software engineering. We need to deliver high-quality software at lower cost and higher velocity to address shifting market trends and stay a step ahead of the digitally native competitors.

What impact has the shift to software engineering had on business operations?

The journey from OEM customer requirements to our production lines, or product engineering to production engineering, has to be seamless because we are working with safety critical domains. We cannot afford to use fragmented systems that have little traceability, where you communicate product information in multiple formats. We need to work in model-based system engineering like UML (unified modeling language) with linkage across the entire software development lifecycle.

Sophisticated infotainment displays for high-end vehicle manufacturers used to take years before they were ready for the mass market. Now, consumers have become much more demanding, which means that our time from innovation to scale has to compress.

How are you meeting those cycle-time compression demands?

We are working collaboratively with our engineering teams and have embraced platform thinking where we no longer think about bill of materials; we think in terms of bill of features. We need to keep these features well-tested and shelf-ready so that when we have an engaged customer, we can quickly assemble and deliver new products with those already tested new features.

We are relying more and more on microservices, which allow us to enhance our physical products with advanced software, reduce our delivery cycle times and create unique customer value. Our teams can now focus on the customer-facing features, like the HMI (human machine interface). They no longer need to put significant engineering energy into already-proven features; instead they can spend their time on value-added features for the customer.

The moment you commit to platform thinking, you have a shot at becoming a more innovative organization. You promote the use of enterprise assets, increase the speed to market for new products, and improve your overall cyber-security posture.

What are some key decisions you’ve made about your platforms?

We have thought carefully about operating systems. As an example, Android, even a year ago, was not widely used in the cockpit electronics space. Today, we are seeing that Android is coming front and center, especially for infotainment. This presents a trade-off question for us: When you run on Android, your system-on-chip costs go up. But your engineering costs go down, because the Android platform contains so many well-tested features.

We are increasingly relying on Docker and Kubernetes orchestration to improve our continuous integration and delivery capabilities. We are leveraging hyper-converged infrastructures with hybrid clouds, and SD-WAN and WAN optimization techniques to bring consistency and resiliency to our platform infrastructure.

When you make the right decisions about your platforms, you don’t really have to figure out what versions of libraries or integrated development environment you have to use when you are engineering a new product. We can keep these prebuilt images in our registry. Engineers no longer have to talk to multiple developers or rely on institutional knowledge; they can spend their time on customer delivery rather than on an IT infrastructure issue.

Our engineers are now increasingly more productive and focused on business outcomes. They are thinking about key performance indicators (KPIs), reuse, and continuous integration and deployment. They know that the moment they check in their code, it will undergo automated testing in a real-time feedback loop.

We’ve discussed Visteon’s innovation in product engineering, but how is IT innovating?

At Visteon, we have gotten good at bringing artificial intelligence into our products. We want to bring that innovation into our IT day-to-day operations as well. We are asking, “Can our infrastructure, which supports 5,000 engineers, be self-healing? If we have a problem in a server or storage area, can we use AI/ML tools to automatically fix it, while also creating a ServiceNow ticket for traceability? Can we bring AI into requirements analysis to learn which requirements tend to generate failures, or are in conflict with each other, or won’t work in a given country? Rather than having a human figure these things out, we want a machine to do the first level of analysis.

We are also changing our KPIs from uptime or latency metrics to track our ability to support collaboration. How have we reduced the number of emails the engineering groups are sending around? How quickly can a new engineer receive a full background of what is happening on the team and with a product? How quickly are we able to detect manufacturing issues? 

I tell my team that in IT, we have a 360-degree view of our business processes. We need to play to that strength.

How does platform thinking change the role of the CIO?

For years, manufacturing and industrial companies have enjoyed protected revenue streams due to long manufacturing life cycles and high fixed and switching costs. So, the barriers to entry for new players were very high.

But now, as products are becoming digital, the advantages that manufacturing companies have always had are fading away. This means that traditional manufacturing companies have to transform into digital companies with a software-centric approach to core design, development, and aftermarket strategy. Manufacturing companies that can make the switch will have the best of both worlds. They’ll have an advantage over newcomers to their markets, because of their traditional customer base, but will still be a software-centric company.

This is where the role of the manufacturing CIO becomes really important. If you can figure how to combine your application development lifecycle management with your product lifecycle management in a holistic way, you can capture your engineering data into model-based formats and build one common, application programming interface-based platform across your product development teams. This is our way to enable digital threads within Visteon.

CIOs who can do this, who can bring the power of the platform, an understanding of market trends, and the ability to collaborate with the product development organization, can be the change agents in transforming their companies.

Copyright © 2020 IDG Communications, Inc.

7 secrets of successful remote IT teams