The Perception team builds the system which learns the spatial-temporal representation and their semantic meanings of the surrounding environment of the autonomously driving vehicle (ADV), i.e., the system that “perceives” the world around the car. We work jointly with downstream teams on the optimization and integration into the Waymo Driver. We conduct our own research to address real-world problems and collaborate with research teams at Alphabet. We have access to millions of miles of driving data from a diverse set of sensors, enabling engineers like you to (1) develop methods for efficiently and continuously learning from large scale real-world data, to (2) develop models and model training at scale, to (3) analyze real-world behavior and develop systems for handling the complexities of interacting with the real-world, and (4) optimize models for our onboard and offboard hardware.

The DRAW team is a perception problem-domain team - our name stands for Degraded Road Surfaces and Weather. We take a high-level business problem like “waymo vehicles need to drive safely in snow” and use whatever technologies and tools we need to solve the problem at hand. Most of our work is ML-related. Recently we have been working with both large supervised multi-model models (lidar+camera+radar) as well as few-shot detection using Vision Language Models (VLMs). We work closely with perception platform teams that build infra for us as well as behavior teams that focus on changing the car’s behavior in response to new outputs we produce.

Previously, our main focus was on driving in rain and dense fog. The team built Waymo’s first ever ML weather estimators to determine the weather around the vehicle and set the vehicle's speed appropriately. We also developed signals to determine when our sensors are in need of cleaning. Finally, we delivered ML models that let the vehicle avoid floods and puddles. We have also worked to make the waymo ADV driver appropriately around potholes, sand, trenches, debris, etc. Our new focus area is snow - we need to build models to understand road friction, snow accumulation, tire tracks in snow, etc.

In this hybrid role, you will report to a Technical Lead Manager.

You will:

• Design and deploy machine learning models that take camera images, lidar data, and radar data and estimate weather conditions
• Work with customer teams to understand what weather-related needs they have
• Build tools and evaluation systems that let the team rapidly launch new weather prediction ML models
  

You have:

• Bachelors in Computer Science, Robotics, Machine Learning or similar technical field of study, or equivalent practical experience
• Experience training and deploying ML models that consume either camera images, lidar data, or radar data
• 4+ years of Industry experience working with large industrial codebases and development tools
• Strong Python experience
• Some experience with C or C++
  

We prefer:

• MS or PhD Degree in Computer Science, Robotics, Machine Learning or a similar discipline with 2+ years of industry experience working with large industrial codebases and development tools
• Deep C++ skills
• Robotics experience