Becoming an engineer in your living room? Udacity makes it possible. (Photo: iStock/alvarez)

How to become a driverless car engineer – online!

Technology and Business

Angelo Rychel

Angelo Rychel



"Our goal is to democratize education"

David Silver, Udacity

Talent shortage is one of the most pressing problems automotive companies are facing when it comes to developing self-driving cars. Now online education company Udacity has created an online curriculum that could fundamentally change the way talents are trained and recruited. We spoke to David Silver.

Despite not being old enough to drive, 15-year-old British high school kid Mikel Bober-Irizar is already capable of programming self-driving cars. “He is one of our most successful students,” says David Silver, who is leading the program on how to become a self-driving car engineer. Although a high performer among his fellow students, Mikel has yet to see a lecture hall from the inside. He is currently studying from home – online.  


Mikel’s case is pretty remarkable, which makes sense considering he is part of a remarkable project – one that epitomizes the car industry’s rapidly increasing demand for software engineers. Be it large OEMs, suppliers or mobility providers like Uber or Lyft, they are all targeting the small pool of talents with programming skills for self-driving cars. And that’s not all: they are also competing with Google, Microsoft, Amazon and countless other companies that are searching for robotics specialists.




At the same time, universities are unable to meet the ever-growing demand. This is why OEMs are acquiring start-ups for unprecedented sums. Or why Uber took on almost all of Carnegie Mellon University’s robotics faculty to launch its autonomous car program. In this contested environment, automotive players are looking for new ways to train and recruit talent. One of them is Udacity’s Self-Driving Car Nanodegree program, a fee-based online credential. Its promise is certainly intriguing: become a driverless car engineer online – in just nine months!

Udacity's self-driving car engineer Nanodegree program started in 2016. (Photo: iStock/Rawpixel)

Udacity is an online education company that offers courses accessible to interested people around the world.  The company is the brainchild of German scientist Sebastian Thrun, an autonomous driving pioneer who won the DARPA Grand Challenge in 2005 and initiated Google’s self-driving car unit. Udacity’s mission is to democratize education through its sophisticated fee-based Nanodegree programs in fields such as artificial intelligence, robotics, machine learning as well as web and mobile development. It was only a question of time until Udacity would also put a focus on driverless cars – and this is where David Silver came into play.


HOW TO BECOME A DRIVERLESS CAR ENGINEER – ONLINE! - 3Silver, a Princeton computer scientist with a Stanford MBA, smiles a lot when he talks about his passion for driverless cars. “They are going to change the world in ways we can’t even imagine,” he says, “I’m really excited to see how it makes the world a better place.” Embracing typical Silicon Valley values, Silver shows a deep-seated belief in technological progress combined with a hands-on mentality: if you want to change the world, just do it! Having worked in various software and consulting jobs, he started enrolling in free online courses at Udacity in 2015. “There were courses related to automotive topics or robotics, but no specific one on self-driving cars yet. It was all kind of a mishmash, not very well organized,” says Silver. He started writing about his experiences in his blog. Then, one day, Silver received a mail from Sebastian Thrun. “He said: Why don’t you come over to Udacity and put together a course on autonomous vehicles that helps people get hired into the industry?”




According to Silver, the Nanodegree program started as what he refers to as a “flyer”. “We just wanted to see if there is interest in this.” When the course was launched in October 2016, the response was overwhelming: more than 15,000 people applied for just 250 slots. Udacity quickly increased the size of the cohort to 500 students, with a new one starting every month. “Our goal is to democratize education so we want to teach as many people as possible,” says Silver.


Can anyone now become a self-driving car engineer? Well, in theory yes. “We don’t require our students to have a university degree – but many of our students have master’s degrees, PhDs or are even professors,” says Silver. Degree or not, there are several prerequisites that applicants need to fulfill: this includes an intermediate programming ability as well as basic knowledge in physics, algebra and calculus. Many students therefore come from technical fields like computer science, automotive or electric engineering. However, all career changers also get a chance: “The other day I read a review from an accountant who is taking the program. He said it’s tough, but he’s enjoying it,” says Silver laughing.

Udacity's students can test their programming skills on a driverless car. (Photo: Udacity)

Tough – that is an assessment that Silver can subscribe to. “It’s time-consuming. Our students on average spend 20 hours a week on the courses. Most of them are doing it on top of their full time jobs.” The nine-month course is divided into three terms. Each term has a tuition fee of 800 U.S. dollars and consists of various video lessons on specific areas of autonomous driving – be it deep learning, sensor fusion or path planning. To pass each term, the students need to master project tasks. For instance, writing a program that detects highway lanes on a video stream. Or using sensor data to build a map of a car’s environment.


But how do students get practical skills? “Those are in fact very practical skills,” Silver says slightly amused. “One thing that surprises many people: most of the work for autonomous vehicles is done on a computer.” This means you don’t need to be physically near the car you are programming. Still, at the end of the program, students will have the chance to test their codes on a self-driving car at Udacity – giving them immediate feedback on how their algorithms work on the road.




The industry seems nothing short of thrilled by the program. Udacity works with renowned companies like Mercedes, BMW, Nvidia and Continental’s Elektrobit, referred to as “hiring partners” that fast track Udacity graduates for job consideration at their companies. By working together with companies in some of the lessons, Udacity students also gain access to an extended job board. If students indicate interest in a position with a hiring partner, Udacity connects them with the companies. But Silver is quick to emphasize that this does not mean exclusive access to the talents: “Our goal is to get our students hired into the industry. We definitely encourage our students to apply for a job if they are excited about it – whether it’s at a hiring partner or somewhere else.”


And the plan works. Although the first cohort of the program will only finish in September, many students were already hired by large companies at the beginning of their studies. Their efforts during the program will be rewarded: according to Udacity, self-driving car engineers can currently earn up to 265,000 U.S. dollars a year. Will Mikel Bober-Irizar be one of them? It’s possible. But it doesn’t seem like he is in it for the money. As he told Automotive News recently: “I want to do something that helps the world.”  


<div id="hs_cos_wrapper_Engage_" class="hs_cos_wrapper hs_cos_wrapper_widget hs_cos_wrapper_type_inline_text" style="" data-hs-cos-general-type="widget" data-hs-cos-type="inline_text" data-hs-cos-field="submit_your_story.icon_text">Letter</div>
Submit your story

Become part of our automated revolution and submit your stories, images and videos

Submit your story
Stay informed

Stay up to speed with our mailing list. Enjoy automated driving content direct to your inbox

Join our weekly briefing
Connect with us

Follow us on our social networks for up to date information and thoughts on automated driving