Software Engineer, Planner Architecture

Aurora Innovation•Mountain View, CA

13d

About The Position

Aurora hires talented people with diverse backgrounds who are ready to help build a transportation ecosystem that will make our roads safer, get crucial goods where they need to go, and make mobility more efficient and accessible for all. We’re searching for a Software Engineer to join the Behavior Architecture Team to develop motion planning projects in areas including, but not limited to, algorithm integration frameworks, fault management, engineering and learned model frameworks, state management and bootstrapping. The Behavior Architecture team is responsible for the software that Aurora’s core motion planning algorithms integrate into; team scope includes integration into the on-vehicle software framework, latency monitoring and architecture optimization, software quality, internal and external interfaces & data structures for the Planner, world representation, route planning, fault management and specialized tools for MP developers. In this role you will: Collaborate within the Autonomy Integration group and with stakeholders across various autonomy subsystems to drive cross-functional design and implementation. Develop software that directly influences the on-road behavior of autonomous vehicles, including architecting the framework and interfaces that core Motion Planning algorithms integrate into. Build onboard infrastructure for detecting and capturing interesting scenarios and anomalous behaviors to accelerate the improvement of the autonomy stack. Enhance Planner robustness and maintainability by reducing faults and improving the flexibility, composability, and testability of the system. Architect and optimize code paths to minimize latency in safety-critical components, ensuring the Planner meets real-time constraints. Improve the Planner’s world representation and data encoding to support both learned and engineered planning approaches.

Requirements

Expertise in Modern C++, specifically for latency-sensitive and safety-critical applications.
2+ years of applied industry experience in software engineering, with a focus on robotics or distributed/asynchronous systems.
Ability to thrive in a highly cross-functional environment, with the technical versatility to contribute across diverse domains ranging from software architecture and algorithms to data analysis and system validation.
Solid understanding of software engineering best practices, including architecture design, code maintainability, test-driven development, and requirements verification.
Demonstrated ability to execute independently: estimating work accurately, delivering on schedule, and communicating blockers effectively.
Proficiency in Python for scripting, data analysis, and automation.

Nice To Haves

An advanced degree in Robotics, Computer Science, or related fields, or equivalent professional experience in autonomous vehicles or robotics domains.
Familiarity with Motion Planning algorithms or control systems within automotive, aerospace, robotics, or similar domains.
Experience working in a regulated, safety-critical environment utilizing systems engineering or requirements-driven development methodologies.

Responsibilities

Collaborate within the Autonomy Integration group and with stakeholders across various autonomy subsystems to drive cross-functional design and implementation.
Develop software that directly influences the on-road behavior of autonomous vehicles, including architecting the framework and interfaces that core Motion Planning algorithms integrate into.
Build onboard infrastructure for detecting and capturing interesting scenarios and anomalous behaviors to accelerate the improvement of the autonomy stack.
Enhance Planner robustness and maintainability by reducing faults and improving the flexibility, composability, and testability of the system.
Architect and optimize code paths to minimize latency in safety-critical components, ensuring the Planner meets real-time constraints.
Improve the Planner’s world representation and data encoding to support both learned and engineered planning approaches.