Applying Machine Learning & Sensing Technologies to Accelerate the Connected Car

"We as an industry have to figure out how to get gigabytes of data off our cars"

Renovo, says the traditional concept of a vehicle platform faces potential radical change in an era dominated by software, electronics and the rise of autonomous vehicle technology.

At the Machine Intelligence in Autonomous Vehicles Summit on 23-24 March in San Francisco, Ioannis Petousis, Head of Data Science at Renovo will be presenting on a panel discussion exploring how we can apply machine learning and sensing technologies to accelerate the connected car.

What started your work in autonomous vehicles?

Autonomous vehicles are a huge technological endeavor that combines some of the latest advancements from various fields - from mechanical engineering and materials science, to electrical engineering and computer science. I love both cars and technology, so it was an easy choice for me.

What are the key factors that have enabled recent advancements in autonomous vehicles?

The recent advancements in artificial intelligence were catalytic in enabling autonomous driving technology, while progress in computer processors allowed those algorithms to be deployed in cars. In addition, the increasing energy density and falling prices of electric batteries has made full electrification possible, which, while not fundamentally necessary, is another major piece in the puzzle for smart, energy efficient, autonomous vehicles.

What are the key challenges to progressing autonomous vehicles?

We have to ensure that self-driving cars can be safely integrated in the existing infrastructure. I believe the key challenges will include the safe operation of the vehicle at all times, security, and the development of an appropriate legal framework. Additionally, the management of autonomous fleets is not a trivial problem and, given that self-driving cars are expected to generate data on the order of 10TB/day, it will involve the efficient transfer and analysis of a large volume of information. While the development of advanced communication networks can help, I believe machine learning will play a decisive role in making the enormous amounts of data produced by autonomous cars manageable.

What developments can we expect to see in autonomous vehicles in the next 5 years?

The next 5 years will be exciting for autonomous vehicles, though how much so will depend on the global political and economic landscape. My hope is that by 2022, we will have experienced successful trials of autonomous taxi fleets in various cities around the world. Perhaps we will also have seen the first commercial vehicles with Level 4 autonomy hitting the markets.

Outside of your own field, what area of machine learning advancements excites you most?

I find the application of machine learning on genetics research fascinating.

The summit will be running alongside the Machine Intelligence Summit, allowing attendees to enjoy additional sessions and networking opportunities to further expand their knowledge.

The Machine Intelligence in Autonomous Vehicles Summit will also be hosted in Amsterdam on 28-29 June, alongside the Machine Intelligence Summit. Early Bird Passed end 10 Feb - don't miss out on this big saving. Register here.