Comment: How AI, data, and software are transforming automotive safety

Imagine a world where the fear of road accidents is a distant memory. A world where cars intricately understand their surroundings, can see far into the distance even in complete darkness, and sense danger before it happens. Cars that can predict future events in a fraction of a second, seamlessly communicating with other vehicles, infrastructure and humans.

Here at Volvo Cars, we’ve been envisioning a safer future on the roads for nearly 100 years. Because if we can’t imagine the future, we can’t build it.

I’m excited to see new technological advancements making roads and cars safer than ever before. These innovations are ushering in a new era of safety. Let me explain.

The history of automotive safety
Back in the 1950s, automotive safety innovations were passive by design – they protected passengers when an accident happened. Safety belts, airbags and strengthened safety cages are all great examples of proactive safety innovations that have saved lives around the globe.

In the 1970s, active safety features began to be adopted in cars. These features were developed to help prevent accidents in dangerous situations by assisting drivers in detecting and avoiding crashes. Innovations such as Blind Spot Information System (BLIS), adaptive cruise control and anti-lock brakes emerged during this era, significantly improving car safety.

Data makes a big difference
It’s no secret that data has long held the answer on how to improve automotive safety. It has driven numerous safety innovations over the years, not just for Volvo Cars but across the entire industry. The Whiplash Injury Protection System (WHIPS) and Side Impact Protection System (SIPS) are good examples of our safety innovations inspired by data collected from crash sites since the 1970s.

But in the past, collecting the necessary data hasn’t always been easy. Manual data collection is labour-intensive, and it can be time intensive to contextualise and turn into insight, even with the most fine-tuned, consistent processes in place. Fortunately, new technological advancements have helped to address these issues.

Core compute has changed the game
New cars today are more like super computers on wheels. They have highly sophisticated computing architectures connected to advanced sensors that help understand what is happening in and around the car better and faster than ever before. For context, according to industry analysis, a typical new car today has more than 150 million lines of code and processes 25GB of data per hour. That’s 1,000 times more lines of code than Apollo 11, which landed on the moon.

For car manufacturers that have begun their software-defined journey, the data collected by core compute architectures is invaluable. It allows us to contextualise this data, unlock new learnings, train future safety models and roll them out globally. For Volvo Cars, this is the real value of the software-defined car.

The transition to the software-defined car running on a core compute system is like adding a superpower of real-world, real-time insight and the ability to write new instructions back to the product to respond even better the next time around. It creates a future where cars can start to learn together as one, growing smarter from each other’s individual experiences.

Becoming software defined
The transformation I’ve described above is perhaps one of the most challenging that any car manufacturer will ever undertake. It represents an even greater milestone than the transition to electrification.

In addition to the innovations mentioned, it requires a completely new organisational infrastructure. This includes creating a software layer than can run across all cars and connect seamlessly with all partner and smart city ecosystems. It requires the most advanced connectivity to enable car communication, cybersecurity protocols to ensure safety, edge computing to reduce latency and a data warehouse to power machine learning.

It requires a complete change in mindset and culture, thinking software-first. It requires the bravery to journey into the unknown and the ability to convince leaders that this is the correct path to take. Success is about continuous innovation, hard work and collaboration, empowering software engineers to make meaningful change in the world by combining energy, matter and code in new ways.

And finally, it requires vision and purpose, just like the one I began this piece with. Success in the modern world isn’t about using technology as a status symbol. Nor is it about embracing innovations that always accelerate the pace of life into a blur or contribute towards an unsustainable way of living (for the planet or for our drivers). When used to its full potential, technology has a much more important purpose – it can save lives and elevate the quality of life.

Technology with purpose
At Volvo Cars, we have the vision, purpose, technology, heritage and roadmap. Our human-centric approach to technology means we use it to make meaningful change in the world – improving safety and elevating the quality of life for our drivers. We’ve been relentlessly using data to improve safety while embracing new technologies that will allow us to raise the bar further.

Thanks to our recent technological transformation, we can harness the power of AI, data and software to build safer cars more efficiently than ever before, bringing us closer to our ambition of a zero-collision future.

Alwin Bakkenes, head of Global Software Engineering, Volvo Cars