Sr Robotics Engineer- Planning and Autonomy

Andromeda Robotics•San Francisco, CA

2d•$200,000 - $240,000

About The Position

Andromeda is hiring a founding member of its Mission Autonomy team to create the long-duration decision-making layer that orchestrates the robot Abi’s activities across an eight-hour shift. Mission Autonomy will continuously decide what Abi should do next, balance task-level trade-offs, and adapt plans while coordinating existing capabilities such as conversation, waypoint navigation, obstacle avoidance, perception, gestures, and operator tooling. The engineer will propose and build the Mission Autonomy architecture from the ground up, using frontier models and simulation and real-world evaluation to train and launch the system. The role combines research and practical robotics, requiring architectural leadership alongside hands-on prototyping and implementation. The goal is for Abi to navigate homes, decide when to join interactions, maintain her character, and recover appropriately when plans fail. The role will establish a path from simulation prototypes to robot demonstrations, including configuration, observability, test gates, and rollback mechanisms. It will also establish evaluation protocols and reporting to measure long-horizon mission performance, required interventions, and limiting failure modes. The first three months focus on understanding Abi’s real-world operations, defining the architecture and six-month plan, building and evaluating a working simulation slice, and taking the prototype onto a robot in the lab.

Requirements

At least 5 years of experience building autonomous systems.
An MS or PhD in Robotics, Computer Science, or a related field, or equivalent practical experience.
Autonomy experience in self-driving, mobile robotics, warehouse robotics, field robotics, or a similar area, including at least one deployment lifecycle from concept through deployment and operations.
Experience with ROS 2, Nav2, and simulation tools such as Isaac Sim, Gazebo, or MuJoCo.
Awareness of safety-critical design principles for robots operating around people.
Strong software engineering skills in Python, C++, or Rust.
Ability to work across the stack, from high-level planning and behavior to low-level robot interfaces and runtime constraints.
Practical robotics experience, including bringing up hardware, debugging sensors and actuators, and working hands-on with robots.
Strong communication and influence skills, particularly across Australian teams and with engineers whose systems are not directly owned by the role.
Ability to steer system design, define interfaces and contracts, and write initial versions of primitives.

Nice To Haves

Experience building mission autonomy systems in robotics, defense technology, or similar domains.
Research experience applying LLM techniques to long-duration robotic task planning, orchestration, tool use, or error recovery; master’s or PhD-level research is ideally suited.
Experience designing simulation-based evaluation systems, including benchmarks, scenario harnesses, metrics, failure taxonomies, and reporting for long-horizon missions.
Experience with robot planning and behavior architectures such as PDDL, task-and-motion planning, behavior trees, or finite-state machines for baselining.
Technical judgment to reason explicitly about trade-offs and earn the trust of senior autonomy engineers and teams that own adjacent systems.

Responsibilities

Design the Mission Autonomy system architecture and roadmap, challenging initial concepts as real-world requirements become clearer.
Prototype a mission autonomy loop in simulation early to inform architectural decisions before teams commit to an approach.
Co-design and build simulation capabilities with the simulation engineer.
Own technical direction and lead design conversations with autonomy, chatbot, perception, audio, and platform teams; define interfaces and contracts those teams can implement.
Specify the action primitives needed by Mission Autonomy and ensure they are built, either by influencing the owning team or implementing them directly.
Define safe, useful, and understandable degraded behavior when audio, perception, connectivity, navigation, or platform health deteriorates, while preserving Abi’s character.
Establish repeatable deployment patterns from simulation to robot demonstrations, including configuration, observability, test gates, and rollback mechanisms.
Build evaluation and benchmarking systems to measure long-horizon mission performance, intervention requirements, and failure modes.
Learn about Abi’s real-world operation, review and challenge the Concept of Operations, and spend time with the robot in an aged-care home.
Deliver an initial architecture and six-month build plan, including decisions about what not to build.
Build and evaluate a simulation scenario in which Abi navigates to a resident’s room and starts a conversation, testing cases such as an absent resident, route changes, and subsystem degradation.
Take the prototype from simulation to Abi in the lab and establish deployment, monitoring, replay, and evaluation patterns.