Why I joined Udacity to work on self driving cars

First week on the job

This was my first week working on Udacity’s self-driving car team. I’m incredibly excited to continue the work started by Oliver Cameron and MacCallister Higgins on open sourcing self-driving technology.

(All of my views in this post are my own and don’t reflect the views of Udacity, Auro, or anyone else.)

Why I left Auro

Testing the new vehicle in my last week at Auro

Since December, I’ve been helping the awesome team at Auro build self-driving electric shuttles as a software engineering contractor. They gave me a shot to work on this technology when all I had was basic programming skill and a ton of enthusiasm (but very little robotics experience). In return, I was willing to work for a low salary on whatever needed to be done. This ranged from improving their mobile app to writing a ROS node for obstacle simulation. My time at Auro was an incredible experience. The team was super supportive, taught me a ton, and gave me the opportunity to work on really interesting technology. My goal was always to get a full-time position working on self-driving technology and when Udacity gave me that opportunity, I jumped on it.

My interview process

The view at one of my only interviews outside of Silicon Valley

When I felt that I’d learned enough from Auro and the Udacity Self-Driving Car Engineer Nanodegree to contribute as a full-time engineer, I started talking to hiring managers and applying to companies that work on self-driving technology. In total, I applied to 14 companies, interviewed with 11 of them, and got offers from 3 of them. The whole process took me 7 weeks.

I can’t go into the specifics of the interviews because of the NDAs, but I can say that everyone was super friendly and pretty reasonable during the interview process. Interviews ranged from extremely informal discussions of my experience and how I’d want to contribute in the role to highly technical, academic questions. A few small companies gave very involved take-home assignments.

For several companies, my lack of theoretical depth in concepts like convolution or Kalman filters seemed to be an issue. Others admitted midway through the interview process that they only wanted engineers with Phds or 5+ years of experience. In a few cases, I decided that the company wasn’t a good fit for what I wanted to do and dropped out of their process entirely. It was a very challenging time and I often struggled to balance interviewing with my work at Auro and my course-load. I’m glad that I made it through and found such a great fit with Udacity.

Why I chose Udacity

Udacity Intersect Conference

Udacity has been an integral part of my journey with self-driving cars since the beginning. Taking their free AI for Robotics course in January 2016 is what first gave me the confidence to pursue this dream of work on autonomous technology. Reading the story of David Silver’s path to working in the field inspired me as well. He gave me lots of good advice when I first moved out to Silicon Valley. The chance to work with him and the other good folks that I’d met through the Nanodegree was a huge draw.

Udacity’s work on open sourcing self-driving technology is very important. The Autoware contributors have made great strides in bringing openness to the very secretive (and often combative) world of self-driving cars. There is still a great deal of work left to improve the quality of open source software in this domain. The most exciting thing about coming to Udacity is the opportunity to work on this mission alongside other students, hiring partners, and open source contributors. Udacity has the resources to let me focus on building great autonomous software and organizing development communities without stressing about acquiring hardware or raising funding.

What’s next?

My hope is that the current collection of Udacity challenge results is just the beginning. Starting with the Didi Challenge and continuing on into testing the code of contributors on a physical car, I want to provide anyone passionate about autonomous vehicles with the resources and community to really contribute to this field and grow their skills at the same time.

Do you have ideas for what this open source effort should look like? Do you want to know more the interview process for autonomous vehicle positions? Hit me up on Twitter or LinkedIn.

Making software and other stuff