Robotics engineer (multi-body dynamics)

Cosmoserve Space · Hyderabad, India

Company Description Cosmoserve Space is dedicated to building a cleaner and more sustainable future for space exploration through advanced active debris removal technologies. The company focuses on high-tech, cost-effective solutions that address the growing challenge of space debris and help safeguard the orbital environment. With a team of passionate experts and a strong engineering culture, Cosmoserve Space is at the forefront of responsible and next-generation space activities. Team members work on impactful projects that ensure the cosmos remains accessible and safe for future missions. Applicants can expect a mission-driven environment that values innovation, collaboration, and technical excellence.

ABOUT THE ROLE

Cosmoserve Space is developing precision robotic systems for space-adjacent environments where dynamic accuracy is mission-critical. As a Robotics Engineer specialising in Multi-body Dynamics (MBD), your primary focus is building, simulating, and validating high-fidelity models of complex robotic systems — capturing joint interactions, contact forces, flexible structures, and actuation behaviour across an entire mechanism. You will work with industry-leading MBD tools and develop simulation environments from the ground up. Dedicated Controls Engineers own the controller design, and Software Engineers handle large-scale software development; your role is to produce the accurate dynamic models and simulation outputs they both depend on.

KEY RESPONSIBILITIES

Multi-body Modelling

•  Develop high-fidelity multi-body dynamic models of robotic systems including manipulators, deployable structures, and contact-rich mechanisms using Mu Jo Co, Adams, Simscape Multibody, or equivalent MBD tools.

•  Define and parameterise joint types, constraint equations, mass/inertia properties, and actuation models to accurately represent physical hardware.

•  Model flexible body effects, structural compliance, and backlash where relevant to system performance.

•  Conduct system identification experiments on real hardware and tune MBD models to match measured dynamics.

Simulation & Analysis

•  Set up and run forward dynamics, inverse dynamics, and quasi-static simulations to analyse robot behaviour across operating conditions.

•  Perform contact and collision simulations — including soft contact, friction, and impact — using Mu Jo Co and complementary tools.

•  Carry out parametric studies, sensitivity analyses, and Monte Carlo simulations to characterise system robustness and performance envelopes.

•  Integrate MBD models with ROS/ROS 2 pipelines to enable hardware-in-the-loop (HIL) and software-in-the-loop (SIL) testing.

Collaboration & Documentation

•  Hand off validated dynamic models and simulation results to the dedicated Controls Engineering team, providing the parameter sets and behavioural data they need to design and tune controllers.

•  Work with the Software Engineering team to expose simulation environments and model interfaces through ROS/ROS 2, scripting lightweight Python or MATLAB wrappers as needed — without owning the larger software codebase.

•  Model straightforward mechanical parts and assemblies in CAD; extract inertial properties and build simulation-ready URDF or Adams representations from these models.

•  Maintain clear model documentation, version-controlled simulation environments, and reproducible experiment workflows.

•  Contribute to technical reports, design reviews, and — where applicable — external publications or conference papers.

REQUIRED SKILLS

MBD Simulation: Mu Jo Co (primary), MSC Adams, Simscape Multibody, Issac Sim

Robotics Middleware: ROS / ROS 2, URDF / XACRO, Gazebo, RViz

Scripting & Analysis: Python (working proficiency), MATLAB / Simulink, Git — for model scripting, data analysis, and automation (large-scale software development handled by Software Engineers)

Dynamics & Mechanics: Rigid body dynamics, Lagrangian & Newton-Euler formulations, Contact mechanics, Flexible body modelling

Control Fundamentals: Working understanding of PID, computed torque, and MPC — sufficient to interpret controller requirements and produce relevant simulation outputs for the Controls team

Math Foundation

Linear algebra, Differential equations, Lie groups / SO(3), SE(3), Numerical integration methods

CAD Modelling

Solid Works — part modelling and assembly (non-complex geometry), mass/inertia extraction, STEP / STL to URDF or Adams pipeline

Analysis & Validation

System identification, Parametric studies, HIL / SIL testing, Data analysis (Num Py, pandas, MATLAB)

NICE TO HAVE

• Experience with NVIDIA Isaac Sim or GPU-accelerated physics simulation environments.

• Familiarity with flexible multibody dynamics (FEM coupling, Abaqus or ANSYS co-simulation).

•  Knowledge of space mechanisms — deployable booms, solar array drives, robotic arms for on-orbit servicing.

• Background in sim-to-real transfer techniques and domain randomisation for MBD models.

• Prior publication, thesis, or project demonstrating MBD modelling of a complex robotic system.

QUALIFICATIONS

• B. Tech / M. Tech in Mechanical Engineering, Robotics, Aerospace Engineering, or a closely related field.

• 2–3 years of hands-on experience with multi-body dynamics simulation in a robotics, aerospace, or mechanical systems context.

•  Demonstrated proficiency with Mu Jo Co and at least one additional MBD platform (Adams, Simscape Multibody, Py Bullet, or Webots).

•  Solid theoretical grounding in Lagrangian mechanics, Newton-Euler dynamics, and constraint-based formulations.

•  Ability to script and automate simulation workflows in Python or MATLAB; large-scale software development is not required.

• CAD proficiency in Solid Works or Fusion 360 — able to independently model straightforward parts and assemblies (non-complex geometry) and prepare them for simulation import, including inertia extraction and joint definitions.

•  Comfortable working alongside dedicated Controls and Software Engineering teams, providing simulation outputs and model interfaces rather than owning those domains.

• Strong numerical skills and comfort working with large, multi-variable dynamic systems.