Magna on the technologies and systems powering the car of the future

Automotive history is littered with examples of failed “car of the future.” They include rotary engines for the mass market, automatic seat belts, and leatherette seats.

Global consulting firm PwC predicts that the car of the future will be “electrified, autonomous, shared, connected and yearly updated.” There is industry-wide consensus that vehicles will be software- and systems-defined.

Vehicle of the future innovations on display at the Magna stand at IAA International included advanced driver assistance systems, a dedicated hybrid transmission system, smarter mirrors for safer driving, scalable battery systems, melt recyclable foam and recycled aluminum alloys.

Automotive Industries (AI) asked Sharath Reddy – SVP Corporate R&D at Magna, how the company balances a system-level approach to technology with customer-centric development.

Reddy:  The first is early engagement of the customer and the second is time to market. The reason our Chinese customers move so fast is that they make decisions in real time. Through this they can reduce a typical four-year program for a new model to 18 to 24 months.

When it comes to technology, the key factors are that it must be modular and scalable. If not, you are in trouble because you are not creating a value proposition.

A value proposition for an OEM is reducing cost or adding performance. If not, then one must ask why we are doing the innovation in the first place.

Another factor that needs to be addressed is being victims of our own organizations. Companies work in silos such as engineering, purchasing, design, sales, and administration where it is very difficult to cross the boundaries between the different sections.

I point out to customers that Tesla’s success is due to all the different departments working closely together. Many Chinese customers follow the same model.

There is nothing stopping existing OEMs and Tiers from doing the same thing. Western OEMs have the technology and tools to compete against the likes of Tesla and the Chinese. do it better.

Since I took over the position of SVP for corporate R&D, I have been preaching “do not try to change the whole car. Let us take bite-sized chunks. We now have some cross domain projects which demonstrate this approach. When we engage with customers and they see the sample and a video of the innovation, they are, like, “wow!”

They know that the technology they are investing in will not be outdated in two years, but will be scalable into the future. We have meetings with German companies represented by people from various domains at vice president level. After the presentation, they say “we all need to need to work together to source Magna on this product.”

AI: Does ADAS have to be intuitive to be effective?

Reddy:  Anything that is interacting with the user must be intuitive. In Europe, where vehicles start chiming when the speed limit is exceeded, people hate the noise and want to turn it off.

So, what was supposed to be a way to avoid getting a traffic ticket and to increase safety has become a feature where the consumer hates you for it, and turning it off is not very intuitive.

A good example is the Apple iPhone. Blackberry and Nokia were around before them, but their user interface was not very intuitive. We can learn from that industry to make the ADAS features more intuitive.

Another example is the lane keep assist and lane departure warning introduced in North America. If you crossed into a different lane it chimed. The idea was to bring you back into your lane, but a less annoying alternative to a chime is a gentle nudge from the steering wheel.

There have also been vibrating seats, which people did not like because the car was telling them what to do.

If you look at the number of sensors some automakers are adding to their cars one has to ask whether they are meeting customer needs. To help our customers make the right decisions Magna has built a demonstration unit at the Graz plant. Automakers can experience the different functions and decide whether they add value before building a single prototype.

AI: How does Magna help OEMs close the gap between engineering intent and everyday usability?

Reddy:  It comes back to a modular concept. There are big car companies which make the full range from luxury cars to the entry segment. Some of the systems we are developing are not affordable at the lower end. We have seen vehicle prices going up significantly since the pandemic, and that is affecting sales.

There should be a greater focus on scalability and modularity for the entire range.

To compete against the Chinese exports Western companies must change their approach. There should be a closer partnership between the OEMs and suppliers such as Magna.

Both parties benefit. At Magna we have had the most success when there is a strategic relationship at the highest levels of the two companies, starting from the very earliest phases of development.

Trust and partnership are a two-way street. Magna is working on innovations for which we need feedback from customers, but we will only show them to companies which we fully trust, and that is the power of strategic relationships.

AI: Is interior sensing the next frontier in vehicle safety?

Reddy:  We have done a good job on exterior sensing by adding cameras, LiDARs and more.

But when it comes to the interior, we have still not solved the child left behind challenge or driver monitoring. There are fatalities and injuries which could have been prevented had there been driver monitoring in the vehicle.

The fusion of interior cabin radar and cameras presents a tremendous opportunity to make interior sensing better. We are using too many sensors inside the car. Fusion of sensing technologies which has been similar to what has happened outside the car will help improve safety.

AI: How does Magna’s mirror-integrated interior sensing solution contribute to vehicle safety and affordability?

Reddy:  I was one of the key people who made this innovation happen at Magna, and I am very proud of what we have achieved. The typical driver monitoring camera is mounted on the A pillar, the steering column or in the central stack.

Styling is different from platform to platform, and tooling costs vary. It has become a big challenge to design a vehicle to incorporate and assemble these different configurations. When you put the camera in the mirror, you just change the mirror.

What was a simple two-megapixel camera in the steering column is now a five-megapixel camera which covers both the drive and the passenger. We can offer some unique features, such as video conferencing or, for the new generation, taking selfies. Drivers also do not have a camera pointing directly at their face. It is hidden behind the mirror.

Another feature we are looking at is to use the camera to measure the blood alcohol level of the driver using spectrometry.

It resonates with automakers because the manufacturing and assembly costs are lower. Within Magna, it has brough our mirrors group MML and Magna electronics closer together.

AI: Where does SDVerse fit into the Magna philosophy of software development?

Reddy:  The transition to software defined vehicles creates opportunities for the sharing of software features. We have seen this in the cellphone industry, and we think it is the way to go.

As an industry we are spending too much money developing the same software again and again. It is time for us to standardize the hardware so OEMs can use the application layer to differentiate their models and ranges.

Unfortunately, some of the automakers have struggled implementing software defined platforms. They are now reaching out to Magna to partner and collaborate with them.

I think this is the enabler for the future of automotive industry, where the brand image of the car company can be safeguarded with the unique features that they want to offer while reducing the cost of development and manufacturing.

Given the trend, we founded SDVerse together with General Motors and Wipro. It is a marketplace where you can select software features. Even companies like General Motors were not aware of some of the software work being done by Tiers. Domain.

SDVerse gives OEMs and Tiers the opportunity to say, “there are five companies doing the software feature, so now I can do software sourcing separately from hardware source.” That trend is going to accelerate. Most recently Stellantis decided to partner with SDVerse to access and commercialize automotive software.

AI: How does Magna approach the use of AI in automotive applications?

Reddy:  We are using AI to help transform the company. AI is completely transforming the way we visualize and develop new features through virtual features and virtual validation. Several OEMs have approached us to increase levels of robotic automation, which are only possible with AI.

AI helps us to be more efficient, and we use it for developing and testing features for our customers.

In the manufacturing plants you will find it in a simple robotic arm for picking and placing, visual inspection and in digital twin planning for factory automation.

For software development we are using AI for new large language models (LLM) and vision language models (VLM) by incorporating them into our software tools such as GitHub. The result has been a 20 to 30% efficiency improvement in software development.

AI: Are affordability and time-to-market as important as innovation to Magna?

Reddy:  They are equally important. We cannot spend three to four years launching a new project and then two to three years developing the innovation. Five to seven years from concept to market is way too long.

Our approach with every innovation is to determine whether it resonates with the customer, which would be the OEM. Then, in partnership with the customer, we identify two or three features which are strategically important. Through a combination of partnerships and AI enablers, we can reduce the time to market. The key is communication.

AI: What is Magna’s perspective on the future of automation and workforce evolution within manufacturing?

Reddy:  I think there is a lot of hype in industry right now about how robots are going to replace humans in the manufacturing plants.

We are taking a phased approach. The first priority is safety of the workers. Some of the tasks in a manufacturing plant require heavy lifting, and we look for ways to help the operator. Automation also improves quality through greater reliability and consistency.

You can see this in our collaborative robots or cobots, where the robot helps the human. Human-machine collaboration addresses some of the challenges of hiring operators. We need to think differently. How do we support the new generation of employees with automation?

Using the collaboration approach I think we can handle the transition in a seamless way.

An example is the visual inspection of seats. Traditionally you had a human who would spot a wrinkle on the seat and a second operator to iron it out. Humans occasionally miss wrinkles, and that becomes a quality issue for the automaker.

Our vision inspection system identifies the folds or wrinkles, which are dealt with by a human operator. We started with two cameras and now have a 78-camera system to scan the whole seat. Cameras are getting cheaper and are helping to improve quality and reliability – which means a happier customer.

AI: What is next for Magna?

Reddy:  We have a very promising future, with plants in North America, Europe and China serving a very wide customer base with a very large product portfolio.

Internally, we are focusing on breaking down barriers between the domains in the company through cross development of products. This is an area in which I think we can help our automakers customers improve. There are tremendous benefits to working across the different domains within a company.

One of the breakthroughs we have developed is a way to save up to eight kilograms on a battery. Another is simplifying harness assembly, which at present is largely done manually. If you can simplify the harness, you can reduce material costs and make it easier to automate the assembly.

We are also using technology to improve latches, door structure, and seating to make it easier to access a car. Here again we see the advantages of working across domains.

When you work in silos, the individual components may all meet specifications and be well designed, but when you put them together, customers point out problems which were not identified during the design phase.

When OEMs and Magna think differently and work across domains, we can save costs and time to market.

Our second focus is on factory automation. We have 341 manufacturing plants and a huge number of employees. By automative some tasks we can make Magna products even better and reduce our cost structure while making our manufacturing processes safer for employees.

These are incremental steps. We do not need or want to change the whole world, or how the industry designs cars.