Senior/Staff Software Engineer, Behavior Fault Architecture and Detection Validation

Nuro•Mountain View, CA

About The Position

About The Team The Systems Engineering team is responsible for the requirements, architecture, and validation of autonomous driving capabilities across engineering disciplines. This includes designing performance metrics, evaluation methods, and criteria for success, which the team then drives cross-functionally via requirement definition and system validation. Systems Engineering works at the intersection of hardware, software, and robot operations, with a deep understanding of technologies in all three. We are a small, high-impact team that sets the checkpoints for autonomy deployment. About The Role As a Senior/Staff Software Engineer, Behavior Fault Architecture and Detection Validation you will work very closely with the software engineering, autonomy product, and data science teams. You will partner closely with our software engineering team to identify faults within a complex, autonomous software stack; refine fault handling criteria; define interface requirements; design the fault and fail-operational architecture; implement new test methodologies to support verification and validation; and contribute to risk quantification. As a key part of Nuro's systems approach, you will contribute to maintaining Autonomy’s fault-centric architecture and stress test the autonomy software stack from system level down to modules through deterministic and probabilistic software-driven techniques towards a fail-operational system. Your work is expected to have short term impact in our next deployments, as well as long term impact on autonomy feature roadmapping and Nuro’s scaled fleet operations. About the Work Dive into the system-level implications of Nuro’s autonomy architecture and software to understand inter-dependency of failure modes for Behavior and Routing. Define requirements for fault handling throughout the autonomy software architecture at the interface between and within software components. Drive fault-centric validation concepts from prototype to production to test. Support the development of tools and data engineering frameworks that can inject faults into the autonomy stack and interpret results from curated amounts of simulation, on-road, closed course testing to assess mission capabilities and safety. Work closely with autonomy developers to extract high-impact metrics and performance indicators that will test our robots’ readiness to drive autonomously on public roads in the event of degraded states of software. Contribute to tooling for automated evaluation of issues in test logs that check for satisfaction of contingency requirements.

Requirements

3+ years of technical work experience in robotics SW or relevant area.
Bachelor’s Degree or Master’s Degree candidate in Computer Science, Math, Electrical Engineering, Mechanical Engineering, Robotics, Physics, or related field.
Strong programming and algorithmic problem solving skills in C++/C.
Technically fluent and experienced working in a team with strong software engineering practices.
Highly collaborative in nature with strong abilities to think and communicate analytically and effectively.
Self starter and fast learner - you should be passionate about picking up new skills and approaching unstructured problems from first principles.
You must have at least 3 years of experience in multiple of the following (not necessarily in industry): Software development in robotics Behavior, motion planning, or routing.
Experience with Linux or Unix computer systems engineering such as communication protocols, designing fault-tolerant systems, and atomicity and coordination of concurrent activities for complex autonomous software.
Analyzing and managing FDIR (fault detection, isolation, recovery) for the software stack on semi-autonomous or autonomous vehicles or equipment.
Defining requirements using data-driven and first-principle techniques.
Implementing test coverage for faults.
Identifying emergent behavior in a safety-critical complex system including interactions between modules.
End-to-end functional decomposition, defining physical/logical architectures, allocating functionality, and reliability analysis of safety-critical complex systems via STPA or other hazard analysis techniques.
Adaptive stress testing or probabilistic software diagnostics.

Responsibilities

Dive into the system-level implications of Nuro’s autonomy architecture and software to understand inter-dependency of failure modes for Behavior and Routing.
Define requirements for fault handling throughout the autonomy software architecture at the interface between and within software components.
Drive fault-centric validation concepts from prototype to production to test.
Support the development of tools and data engineering frameworks that can inject faults into the autonomy stack and interpret results from curated amounts of simulation, on-road, closed course testing to assess mission capabilities and safety.
Work closely with autonomy developers to extract high-impact metrics and performance indicators that will test our robots’ readiness to drive autonomously on public roads in the event of degraded states of software.
Contribute to tooling for automated evaluation of issues in test logs that check for satisfaction of contingency requirements.