Simulation Realism Engineer (Mod/Sim)

iMETALX•Sausalito, CA

11d•Onsite

About The Position

iMETALX, Inc. is pioneering innovations in space accessibility and sustainability. Our mission is to deliver cutting-edge solutions in space domain awareness (SDA) and in-space servicing, assembly, and manufacturing (ISAM) to both government and commercial sectors. Our focus encompasses spacecraft autonomy, leveraging advanced technologies in perception, controls, and testing in real-world systems. We are building a small, high-impact team working on cross-domain applications that leverage state-of-the-art computer vision, machine learning, and autonomy. This includes both space applications (ISAM & SDA) and terrestrial analogs that use similar technologies. You will be joining at a stage where your ideas and ownership meaningfully shape the technical direction, culture, and impact of the company. As a Simulation Realism Engineer, you will play a crucial role in creating robust and high-fidelity simulations that enhance our autonomy systems, critical for both space operations and terrestrial applications. Role Overview: We’re looking for a Simulation Realism Engineer to take ownership of high-fidelity synthetic scene generation and simulation realism at iMETALX. As a Simulation Realism Engineer at iMETALX, you will focus on designing and implementing state-of-the-art simulation environments that accurately reflect real-world scenarios for our ISAM and SDA missions. You will be responsible for enhancing simulation fidelity, minimizing the sim-to-real gap, enabling more effective testing and validation of autonomy systems, thereby pushing the boundaries of what is achievable in space operations. Your work will involve collaborating closely with perception/ML engineers, autonomy/GNC, and customers to ensure our simulations meet rigorous performance criteria and enable meaningful insights into system behaviors under varied conditions. You will validate simulations through rigorous testing and foster innovations that lead to exciting advancements in our technology. You will own simulation realism as a product: data quality realism metrics validation methodology scenario generation scalability + automation tooling that lets the team generate high-quality datasets quickly and repeatedly

Requirements

BS/MS/PhD in Aerospace, Robotics, CS, Applied Math/Physics, or equivalent experience
3+ years building production-quality systems in: modeling & simulation, rendering/graphics engines, robotics autonomy simulation, synthetic data generation pipelines
Strong programming skills in Python and C++
Experience with modeling environments, vehicles, or sensor systems relevant to autonomous systems.
Experience with at least one simulation/graphics platform: Omniverse, Unreal Engine, Unity, Gazebo, OpenGL, Vulkan
Understanding of simulation methodologies and techniques to create high-fidelity models.
Ability and willingness to work on-site in Sausalito, CA.
U.S. citizenship due to ITAR export-control restrictions; only U.S. citizens are eligible for this position.

Nice To Haves

Experience closing sim-to-real gaps in robotics/autonomy systems
Familiarity with simulation for space missions, unmanned vehicles, or other real-time systems.
Understanding of visual and sensor modeling including LiDAR, cameras, and IMUs.
Experience with hardware-in-the-loop or software-in-the-loop simulation testing.
Experience scaling simulation workloads via GPU, distributed processing, dataset automation

Responsibilities

Build high-fidelity synthetic data systems
Design and implement photorealistic synthetic scene generation pipelines for space missions
Generate scalable datasets robust and impactful training
Close the sim-to-real gap with metrics
Identify the highest-impact realism gaps affecting autonomy performance
Create measurable realism metrics (quantitative vs qualitative)
Build regression tests that prevent simulation fidelity drift over time
Engineer environmental realism
Implement and tune physics-based modeling of the space environment and sensing HW
Scenario generation & infrastructure
Build procedural scenario generation tools for both spacecraft and their missions
Improve simulation performance using GPUs, parallelism, and cloud-scale dataset generation pipelines
Create developer-friendly interfaces so engineers can rapidly test and train
Collaboration
Collaborate with engineers across disciplines to build an impactful tool