Track Talk W3

Testing Self-Driving Cars in a Big Data World

Johan Jonasson

Sofie Lindgren

11:00-11:45 CEST Wednesday 8th June

The competition in the automotive industry is fierce and everybody wants to be the market leader in autonomous driving. AI and machine learning methods are used to train ever-more capable algorithms for sensor software, and the hardware and electronics manufacturers are racing to develop increasingly potent and effective solutions to run the software on. Soon steering wheels will be heading the way of the dinosaurs. Right?

Not so fast. Unsurprisingly, there are a few obstacles still left to overcome.

In order to develop machine learning technology, you need data. Lots of data. But not just any data. it also needs to be representative, varied, and reliable, with storage solutions capable of storing and and serving up several peta bytes for both development and testing. And even then you’re only halfway there.

To be able to deliver a true autonomous driving solution your jumping-off point must be testing. As a supplier in this industry, you need to be able to demonstrate, not only that your product is capable, but more importantly that it is safe. We’re in the business of saving lives, and “good enough” just won’t cut it.

Join us to learn how we are working to continuously improve an industry leading testing methodology for the next generation automotive active safety systems, and how the context and challenges of self-driving cars and Big Data will likely shape the future of many other software testing fields in the future.