Waymo is the self-driving technology company with a mission to make it safe and easy for people and things to move around. Waymo has deployed a fleet of autonomous vehicles that generate vast amounts of data in a variety of sensing modalities. The perception team at Waymo is responsible for the broad recognition problems autonomous vehicles face. In this capacity, our team builds systems and models, that use the acquired sensor data, to recognize objects, and power our fleet of self-driving cars.
Perception infrastructure involves both onboard and offboard systems. The onboard perception system runs on the self-driving car. It is an extremely high functionality and high performance real-time system. The onboard perception system processes all the sensor inputs of lidars, cameras, radars, and audio. One core design principle of the perception onboard system is to be ML effective, both in terms of running ML models onboard with very low latency, as well as integrating all the ML models into a cohesive system.
The offboard perception infrastructure supports engineering development of the onboard system. It sits on the interface of machine learning and systems. It supports our expansive ML development including data extraction, evaluation, debugging / visualization, data mining, etc.. It also supports systems development including testing, performance analysis, deployment, etc. Waymo has a massive amount of sensor data logged from real-world driving, which is key to our engineering development. Robust and scalable perception offboard infrastructure is at the core of engineering excellence.
In this role, depending on preference and role fit, you will
Work on the onboard perception system, where you’ll get to build an overall picture of the entire perception stack, conduct latency measurements and analysis, optimize system performance, make software design changes as needed, etc. Or,
Work on the offboard perception system, where you’ll learn the nitty-gritty details of the infrastructure to build ML models that actually works in the real world, design and implement ML/perception infrastructure to enable Waymo scale to many more cities and thousands of cars, directly interact with and enable many ML developers to achieve cutting-edge self-driving perception, etc. Or,
Work on both of the above.
We’d like you to have:
Strong in C++ and system design.
Excited about building solutions to complex problems.
Willingness to get your hands dirty and go full stack if need be.
Experience in robotics or machine learning.
Ideally, you also have:
Experience in deploying a large production system with many ML models.
Experience in optimizing high-performance systems.