Autonomous Vehicle Industry Races to Fill Big Engineering Talent Gap
Dec 5, 2018
by Jean Thilmany ASME.org
Transportation in the next 10 years will change dramatically, shifting from automobile ownership to a much greater use of shared mobility options. Many of those rides will be electric and automated.
Once automated or self-driving vehicles (AVs) become commonplace, they’ll turn the entire transportation system on its head, said Tim Papandreou, founder of City Innovate, a San Francisco-based company that uses data to guide city governments on technologies. City designs, trucking and shipping, and mobility management are just a few things that local governments will face during a major transformation.
The engineering job market is another area that local governments will need to tackle. Today, the AV industry is in dire need of engineers capable of designing and working on these types of vehicles, and the systems and infrastructure needed to support them.
“There’s a tremendous talent gap for AV engineers,” says David Silver, head of the School of Autonomous Systems at Udacity, a higher-education start-up of Mountain View, Calif.
Silver isn’t alone. Sebastian Thrun founded Google’s self-driving car team in 2007 and has seen a large number of AV-related start-ups sprout since then. But engineering programs that train the engineers to create and work on AVs haven’t, and are almost nonexistent, he says.
For You: System Helps Self-driving Cars See in Fog AV companies are desperate for talent, he says. But the skill set needed to build self-driving vehicles, and the infrastructure needed to support them requires multidisciplinary skills that most engineering students lack, he says.
Thrun estimated that the overall U.S. tech industry needs about 5,000 engineers. That demand is why AV engineers’ salaries average $240,000 annually compared to an average of $110,000 for software engineers, he said.
“Just that difference in compensation shows how hard it is for AV companies to find good engineers,” Silver said.
Thrun co-founded Udacity, an online school that offers mini-courses and certificates within a number of engineering and computer subjects, about seven years ago. The school established its AV engineering course two years ago. Since then, about 12,000 students have enrolled in the school’s self-driving program to earn a “nanodegree in the field.
AV training “lives at the intersection of math, mechanical engineering, and electrical engineering,” Silver said. “Plus, they need to know computer-vision, robotics, and motion planning, and sensor-fusion, and LIDAR, and RADAR technologies to augment their understanding of this work. That’s part of why it’s so hard to find people for this,” Silver says. “It’s because it’s so hard to find people with a background in all those areas.”
Most Udacity students have engineering degrees and many already work as engineers. But they seek to augment their training and job skills with the six-month AV training program, which offers skills training to help students become job ready. While Udacity’s program isnot formally accredited, the program aims to equip students with the specific skills that match those of the company that the students want to work for.
The first 2,000 AV engineering students will graduate from the Udacity program at the end of 2018. About 100 of them already have job offers, Silver said. “Employers weren’t even waiting until they’d graduated to recruit them,” he added.
“Over time we want to supply industry withas many engineers as needed,” Silver says. “We can scale to as many students around the world as want to take it. They don’t need to come to Mountain View. They have deadlines for their projects and coding exercises and they earn a nanodegree certificate.”
Udacity might be onto something.
Any society looking to expand the use of AVs needs more teachers in civil, mechanical, and electrical engineering, says Max Donath, director of the Intelligent Transportation Systems Institute at the University of Minnesota.
While many U.S. universities and colleges have autonomous vehicle research, projects, and related classes, very few of them have a defined AV undergraduate degree.
The University of Virginia, for example, offers the course “Autonomous Racing: Principles of Perception, Planning, and Control” taught by Madhur Behl, an assistant professor of computer science and systems and information engineering. Students in the course work in teams to build, drive, and race scaled-down AVs. During the process, they learn to use robot operating systems, integrate sensors that include cameras and LIDAR on an embedded computer, and implement algorithms for location determining, mapping, path planning, and control.
“Right now, so many electrical and mechanical engineers don’t have a clue about LIDAR, and they haven’t heard about differential GPS,” Donath said. “We need to rethink our educational environment. We need to get students excited about AV before they move out into the real world.” He suggests that private companies, such as AV manufacturers, team with degree programs to ensure a steady stream of students who graduate with the tools needed for AV engineering. Polaris, the Medina, Minn.-based off-road vehicle manufacturer, for example, might choose to build an AV “so students can see how it works and doesn’t work, and can play around with it,” Donath said.
“Industry will be hiring people who know these are real things and not toys,” he adds. “They’ll have to have employees who appreciate controlled system failures and who understand how LIDAR and GPS works, and the limits of communication within those systems.”
Geographic areas with university degree AV programs may also experience robust growth in AV industries and the money those will bring to their communities, as students may stay in the after graduation and start-ups and related businesses, he said.
Creating a dedicated AV curriculum in higher education might take a bit more time, but the student interest is there, industries are hungry for graduates with AV expertise, and “the salaries are really quite engaging,” Donarth said.
Jean Thilmany is a freelance writer in St. Paul who frequently writes on engineering topics.
AV training “lives at the intersection of math, mechanical engineering, and electrical engineering.” David Silver, Udacity