Waymo is a self-driving technology company with a mission to make it safe and easy for people and things to get where they’re going. Since our start as the Google Self-Driving Car Project in 2009, Waymo has been focused on building the World’s Most Experienced Driver in hopes of improving the world's access to mobility while saving thousands of lives now lost to traffic crashes. Our Waymo Driver powers Waymo One, our autonomous ride-hailing service, as well as Waymo Via, our trucking and local delivery service. To date, Waymo has driven over 20 million miles autonomously on public roads across 25 U.S. cities and conducted over 15 billion miles of simulation testing.
The Planner Evaluation team is working on one of the most challenging problems in autonomous driving: building algorithms to assess the safety and quality of our self-driving cars at scale. Evaluation at scale is of paramount importance in solving autonomous driving: the algorithms we build orient the development of planner software by defining the north star and thus enabling rapid iteration. Almost every improvement to Waymo’s planner is influenced by our work, indirectly (by providing the quality signals for onboard engineers to aim for) or directly (by incorporating our algorithms into our onboard motion planning code). This is no easy feat. As our self-driving car gets better and accumulates more miles, we need rigorous and intelligent methods to automatically assess whether a scene is risky, a maneuver causes discomfort, or an interaction with pedestrians or other drivers could be managed better. This can be done through geometric reasoning, simulation, machine learning, and/or modeling with statistical distributions. Developing evaluation algorithms in self-driving is an uncharted space and thus requires true ingenuity. If you are excited about leveraging a large volume of extremely rich data in an area with lots of unexplored territory to help the advent of self-driving technology, this is the place for you!
In this hybrid role, you will:
We'd like you to have:
It's preferred if you have: