Featured Articles

Future Says S4E4: Taking the Right Approach to Data Analytics in the Automotive Industry

According to Gartner, only 20% of analytics solutions deliver business outcomes. VentureBeat recently reported that a mere 13% of data analytics projects make it to production. Altair’s own international survey data of more than 2,000 professionals found that 42% of respondents said their organization has experienced an artificial intelligence (AI) project failure within the past two years.

If such figures are accurate, it’s clear that many enterprises are failing to realize the full potential of data analytics. This begs a simple question: What’s going wrong?

In season four of Future Says, Altair is turning to experts from throughout the automotive industry for their frontline insights on applying data science and AI. In episode four, we talk with Rajeev Verma, a senior data science specialist at Magna International, to learn more about data science, Industry 4.0, and more.


Magna: The 65-Year-Old Startup Putting Data Science to Work

Magna is a powerhouse of automotive manufacturing. Based in Aurora, Canada, Magna operates hundreds of facilities worldwide, developing and producing a range of world-class mobility systems, components, and complete modules. The company prides itself on its commitment to rapid innovation – so much so that Swamy Kotagiri, Magna’s chief executive officer, often describes the company as a “65-year-old startup.” 

Rajeev Verma joined Magna in 2017 and is part of a team of data scientists working on a varied array of processes. As he puts it, their job is to use data science to empower manufacturing operations. That takes them into areas such as predictive manufacturing, demand forecasting, and assembly line balancing. 

This role means Verma is immersed in the world of “Industry 4.0,” which Wikipedia says is characterized by “the joining of technologies like AI, gene editing, and advanced robotics that blur the lines between the physical, digital, and biological worlds.” As Verma describes it, the manufacturing sector is currently experiencing a fourth industrial revolution. Today, he says, change is fueled by data generated from the computing technology and sensors first integrated into production processes back in the 1990s and 2000s. With the arrival of Industry 4.0, that vast data resource is now being put to work beyond its local domain. As a result, manufacturing enterprises like Magna are undergoing a new, more powerful cycle of business transformation. In short, data analytics in the automotive industry looks a lot different than it used to.

On paper, Industry 4.0 promises a wealth of improvements: better products, lower costs, more automation, and so on. Verma also predicts widespread Internet of Things (IoT) integration and the development of interconnected manufacturing systems that link OEMs to Tier 1 and Tier 2 suppliers. 


Bridging the Gap

In human terms, Industry 4.0 brings the role of data scientist to the fore. Which raises another fundamental question. In the context of manufacturing and data analytics in the automotive industry, what does a good data scientist look like? 

Verma provides a clear explanation. In manufacturing, business revolves around physical systems. To take advantage of tools like machine learning and AI, someone in the team needs to understand both the physical processes involved, and how to connect them with the world of data science. 

In this respect, Verma’s educational journey ticks all the right boxes – allowing him to bridge the gap between physical processes and data science. Highlights include a degree in mechanical engineering and a doctorate in electrical engineering systems from the University of Michigan, and a master’s degree in analytics from the Georgia Institute of Technology. At Magna, he’s seeing more colleagues make that same transition from mechanical or systems engineering to data science. As a result, there’s an ever-growing cohort within the company who can span the distance between physical processes and new data-driven technologies.  


The Foundations of Successful Model Development

Verma underlines the value of this multi-skilled perspective when discussing the development of new data-driven models. The initial priorities should be to define the problem, understand the data, then make a sound business case for any proposed solution. At Magna, most use cases originate on the shop floor. That means effective collaboration between systems specialists and data scientists is key. Ideally, he says, the two disciplines should work side by side. Moreover, in selling a solution to customers, Verma believes that data scientists should always be speaking the language of business, not technology.   

At the other end of the development journey, Verma recognizes that MLOps can also represent a significant hurdle. Managing models is often difficult. This is particularly true if there’s a need to scale: “That’s when it becomes very important – how you version your code, how you version your data, how you version your models, how you version your pipelines,” Verma said. With typical clarity, he admits that this might not be much fun, but that it must be done. In the near future, he expects to see the growth of dedicated MLOps teams across industries, focused on overcoming the specific challenges of model deployment and management.


Human Creativity Will Drive the Future of Manufacturing

Verma welcomes the trend of democratization within data analytics in the automotive industry. And he shares some interesting insights into how it might be shaped by newer, younger generations of employees. As Verma sees it, young people have grown up in a world in which they expect immediate oversight over the things that directly affect their lives. These same expectations will extend to their work, he predicts. In data science, this means providing staff with direct access to the data generated by machines they’re responsible for.  

Of course, new generations of employees will also bring their own skills and flair to their roles. Verma appreciates this and believes that sustaining an individual’s spirit of innovation over the long term is a crucial challenge for any manufacturing organization. Above all, Verma thinks the answer lies in a commitment to continuous education, a commitment that’s shared by both employers and employees. Upskilling, along with a willingness to redefine job roles, will help foster a data-driven culture, Verma said. At Magna, this ethos has led to the creation of the “Magna University.” Developed in conjunction with the University of Toronto, it gives employees the chance to pursue new educational opportunities alongside their day jobs. You could say Verma is a proponent of this initiative – after all, he co-teaches a course on AI and machine learning within Magna University.   



Overall, like any good innovator, Verma is a passionate advocate for the pursuit of new skills. And these skills, he emphasizes, don’t need to be limited to formal education and training. Personally, he finds fresh ideas and inspiration in podcasts such as “The Productivity Show,” and the books of Nassim Nicholas Taleb, notably “Skin in the Game.” In fact, for all the talk of new technologies, it is clear that, for Verma, the future of data science in manufacturing will be powered and shaped by human creativity and collaboration. As he puts it: “Developing a passion for educating yourself and reskilling yourself should be job number one.”

Click here to listen to the full episode with Magna's Rajeev Verma. To learn more about the Future Says series and browse previous episodes, visit

Future Says Season 4 is proudly sponsored by Oracle. Oracle offers integrated suites of applications plus secure, autonomous infrastructure in the Oracle Cloud. For more information about Oracle (NYSE: ORCL), please visit


S4 Previous Recap Articles