Autonomy Algorithm Engineer

LUXUAV · Foetz

Apply for this position and join us in building Europe’s next generation infrastructure.

Our values are rooted in responsibility, readiness, and long term thinking. We believe sovereignty must be designed into systems from the start. Readiness is achieved through capability, not procurement. Autonomy is infrastructure, not a feature. We build with the understanding that modern systems carry long term consequences. That is why we prioritise reliability over novelty, integration over isolation, and sustained capability over short term advantage.

Role Overview

Reports to: Head od Software & AI

You will develop the algorithms that let our UAVs navigate, decide and work together on their own:

state estimation, GPS-denied navigation, guidance and control, mission autonomy and swarm

coordination. The work runs from maths and simulation through to flight testing, for operations in

contested, GPS-degraded and communications-denied environments.

Nobody is an expert in every area below. We are looking for strong fundamentals in one or few of them

and the appetite to grow into the rest. If you meet most of the requirements, we encourage you to

apply.

Key Responsibilities

What you will work on:

Your focus will depend on your strengths; the team covers all five areas together.

Navigation and state estimation

• Design and tune sensor-fusion estimators (Kalman-filter family, factor graphs or similar) that

combine inertial, GNSS, vision, barometric and radio-based measurements into a reliable pose

estimate.

• Build or adapt visual, visual-inertial or LiDAR odometry and SLAM for navigation without GNSS.

• Make navigation robust to sensor dropout, jamming and spoofing: detect degraded conditions

and switch to fallback modes.

Guidance, control and trajectories

• Develop trajectory generation, motion prediction and tracking for dynamic flight in uncertain

environments.

• Contribute to guidance laws, including terminal-phase guidance, and to collision avoidance.

• Interface algorithms with the flight controller (for example PX4 or ArduPilot over MAVLink).

Swarm coordination

• Design coordination algorithms for groups of UAVs: task allocation, formation and spacing, and

cooperative search and surveillance.

• Develop cooperative navigation and shared situational awareness, such as relative localisation

between aircraft and distributed estimation.

• Keep the swarm effective under limited, delayed or lost communications, with decentralised

decision-making and graceful degradation as aircraft or links drop out.

• Deconflict aircraft from each other and from external airspace constraints.

Mission autonomy

• Design mission behaviours such as persistent surveillance, replanning when threats or tasks

change, and safe return under sensor or link loss.

• Choose the right decision method for each problem (rule-based, planning and optimisation, or

learning-based) and compare them honestly before they fly.

• Bound every autonomous behaviour with hard safety limits and a deterministic fallback, so

aircraft behaviour stays predictable and certifiable.

• Integrate geofencing and dynamic no-fly zones in line with EASA U-space requirements.

Integration and validation

• Integrate algorithms with the flight stack and ROS 2 autonomy software, and deploy them on

onboard computers within real-time, power and thermal budgets.

• Validate in simulation and on hardware-in-the-loop rigs before field trials, then analyse flight

data against measurable targets (drift, convergence, robustness) and iterate.

• Document designs and interfaces to support certification.

Across all areas, we use machine learning where it measurably beats a classical baseline, and keep the

classical solution where it does not.

Experience & Skills

What you will need:

• A Master's or PhD in robotics, control, aerospace, electrical engineering, computer science,

physics, mathematics or a related field, or a Bachelor's with equivalent hands-on experience.

• 3+ years building estimation, navigation, guidance, control or autonomy software for real

systems: drones, robots, autonomous vehicles, aerospace, marine or similar. Relevant PhD

research counts.

• A solid grasp of sensor fusion and probabilistic estimation, for example Kalman filters (EKF/UKF),

particle filters or factor graphs.

• Hands-on depth in at least one of: SLAM or visual/LiDAR odometry, GNSS-denied navigation,

guidance and control, motion planning, or multi-robot coordination.

• Strong C++ and Python for algorithm development, with numerical libraries such as Eigen, Ceres,

GTSAM or NumPy/SciPy.

• Strong mathematical foundations: linear algebra, probability, optimisation, and control or

dynamics.

• A working understanding of machine learning concepts, with hands-on implementation

experience in supervised, unsupervised and reinforcement learning (for example in PyTorch or

similar), and the judgement to tell whether a model genuinely generalises.

• Experience taking algorithms from simulation onto real hardware and debugging them with real

sensor data.

• Good engineering habits: version control, code review and testing.

• English at upper-intermediate level or above, and comfort working with hardware, software and

flight-test teams.

Nice to have:

None of these is required; each is a strength we will value.

• Swarm or multi-agent work: task allocation, consensus and distributed estimation, cooperative

localisation, formation control or multi-agent reinforcement learning.

• Communications-aware autonomy: mesh networking, bandwidth-limited data sharing between

aircraft.

• Tightly-coupled visual-inertial odometry and factor-graph optimisation (for example VINS-

Fusion, OpenVINS, ORB-SLAM3, LIO-SAM, GTSAM or g2o).

• PX4 or ArduPilot, MAVLink, and simulation tools such as Gazebo, MATLAB/Simulink or software-

in-the-loop setups.

• ROS 2: package development, transform trees, sensor integration and data pipelines.

• Machine learning with PyTorch or similar, including learning-augmented estimation (learned

inertial odometry, differentiable filters, neural feature matching).

• Reinforcement or imitation learning for robotics (for example PPO or SAC with Gymnasium,

Stable-Baselines3 or Isaac Lab) and sim-to-real transfer.

• Guidance laws such as proportional navigation, or trajectory optimisation for agile flight.

• Embedded deployment on NVIDIA Jetson-class hardware (TensorRT, quantisation) and real-time

operating systems (FreeRTOS, Zephyr, NuttX).

• Radio-based positioning such as ultra-wideband or pseudolites.

• Safety-critical development: DO-178C, MISRA C/C++, STANAG 4671 or EASA guidance on

machine learning in aviation.

• Publications (ICRA, IROS, CoRL, RSS, IEEE T-RO) or open-source contributions in robotics,

estimation or learning.

Eligibility

• National of a NATO member state, or of Australia, Japan, South Korea, New Zealand or Ukraine.

• Clean criminal record.

What guides our decisions?

Our values are rooted in responsibility, readiness, and long term thinking. We believe sovereignty must be designed into systems from the start. Readiness is achieved through capability, not procurement. Autonomy is infrastructure, not a feature. We build with the understanding that modern systems carry long term consequences. That is why we prioritise reliability over novelty, integration over isolation, and sustained capability over short term advantage.

[01]

Design with scale in mind.

We develop systems as part of an ecosystem that is intended to grow.

Scale is not an afterthought, but a design principle that shapes architecture, integration, and evolution.

[02]

Build a solid foundation.

We prioritise core architecture, interoperability, and long term resilience.

A strong foundation enables systems to adapt without fragmentation.

[03]

Practice over theory.

We value learning through application.

Systems are shaped by real use and continuous refinement, not by abstract assumptions.

[04]

Commitment beyond delivery.

We take responsibility for what we build over time.

Commitment means supporting systems throughout their lifecycle and ensuring they remain relevant, secure, and under control.

We are hiring talent across multiple roles, contributing to Europe’s readiness and long term capability.

We are building a multidisciplinary team across engineering, systems architecture, operations, and supporting functions. Open roles reflect the needs of a growing secure and autonomous infrastructure company, where integration, reliability, and long term thinking matter. If you do not see a position that matches your profile, we still encourage proactive applications from people aligned with our mission and values.

Position

Experience

Location

Model

Apply now

More roles at LUXUAV