• Automated Driving Systems
  • Automated Vehicles
  • Conferences and Events

Developer Conferences Can Help to Advance the State of the Art in Automated Driving

Sam Abuelsamid
Mar 07, 2019

Smart Car 3

As both software and computer platforms have become more capable over the past few decades, developers have been able to build amazing products around them, including automotive platforms. The increasing complexity of these systems, however, means it can be a challenge to get started in understanding how to utilize those capabilities. Fortunately, technology companies understand that it is in their own business interest to help developers get up to speed with these platforms, and they have stepped up with events like the Nvidia GPU Technology Conference (GTC). 

In my role as a technology analyst, I attend many conferences throughout the course of each year. These events vary in scope and size, from smaller gatherings like Automotive World’s M:Bility conference where practitioners present their latest results to much larger shows like Apple’s World Wide Developer Conference, Google I/O, Microsoft Build, and GTC. Large or small, conferences provide a great medium for learning about the latest and greatest developments in the areas I cover, which are mainly focused on automated and connected vehicles, mobility services, and advanced propulsion. I get to meet with many of the people working in these fields and discuss what they have learned. 

Tracing the Development of Automated Driving Systems

The modern beginnings of developing automated driving systems can be traced to the launch of the DARPA Grand Challenge in 2003. It was only years after the program concluded that advances in both machine learning and computing power from graphics processors spurred the whole effort forward toward commercialization. 

Software engineering is hardly new to the auto industry, as much of the advancement in fuel efficiency, cleaner emissions, improved performance, and safety since the 1970s can be directly traced to computerizing almost every system in modern vehicles. But the embedded software approaches used by engineers, including myself (I worked on automotive electronic slip control systems for 17 years) and many thousands of others, are not adequate for automated driving. 

Using machine learning and developing deep neural networks requires new skill sets. This is exactly why General Motors is now laying off engineers that do not match these new requirements. While many engineers are experts in their fields, those fields are fading in importance compared to automation, electrification, and connectivity. 

An Opportunity to Learn from Experts

Conferences like GTC provide the opportunity to learn from the experts that created these new technologies as well as from thousands of other developers applying the technology in a vast array of areas. For example, work being done in the field of computer vision and image classification in healthcare can be applied to similar problem sets in automated driving, such as pedestrian detection. 

Validating automated driving systems demands far more testing than can ever reasonably be done on the road, requiring vast amounts of simulation. Efforts in video game development can be applied toward developing automotive simulation tools. Augmented reality is expected to be a key in the user interfaces needed to build user trust in automation. 

Nvidia currently hosts seven GTCs around the world, with the first 2019 event taking place March 17-21 in San Jose, California, which will include more than 500 sessions and tech talks. In a rapidly changing automotive industry, engineers looking to hang onto or expand their role need to keep learning all the time. Events like GTC or similar conferences provide an ideal place to do that. Nvidia has kindly provided a discount code for automotive developers interested in attending GTC Silicon Valley this March. Browse the automotive sessions and then register using the code MCXNAVIGANT for a 20% discount.