aimlroles

Senior Perception Engineer — Aerial Vision, Tracking & 3D Geometry

TwentyFour Industries Open · verified Oct 22, 2026
Munich Full-time Senior AI Research

At a glance

Senior AI Research role at TwentyFour Industries. Munich · full-time.

Pay not stated AI Research salary

RoleAI Research
SenioritySenior
LocationMunich
EmploymentFull-time
PostedAug 5, 2026 · 2mo ago
Role brief

aimlroles summary, based on the employer's posting.

What you'll do

  • Build and maintain C++ components for drone detection, tracking, target positioning and trajectory estimates
  • Improve how targets are associated, tracked and filtered, including uncertainty and detector output fusion
  • Combine camera observations with vehicle and navigation data to estimate targets’ positions and motion
  • Integrate neural-network models into the C++ pipeline using ONNX, TensorRT, CUDA and NVIDIA Jetson
  • Replay flight data to reproduce failures, diagnose causes and validate the perception pipeline

What you bring

  • You have shipped perception, tracking, localization or state-estimation software on real robotic systems
  • You bring strong C++ and production-software skills, including experience with performance-sensitive systems
  • You understand coordinate geometry and probabilistic estimation, including filtering, uncertainty and sensor fusion
  • You can integrate learned models into perception systems and assess their behavior, latency and trade-offs

Who this fits

This senior role suits an engineer ready to take ownership of complex perception systems and work across ML, autonomy, simulation and flight testing. Experience with aerial perception, robotics flight stacks or edge deployment is useful. The role is based in Munich, with remote options and relocation support.

From the employer

We’re looking for a Senior Perception Engineer to co-own and advance the production perception stack for autonomous drones. You’ll develop real-time C++ systems that combine learned visual detections with multi-object tracking, camera geometry, vehicle state, and monocular 3D target estimation. The role is focused on real world and deployable software: translating algorithms and models into maintainable onboard components, validating them on recorded and live flight data, and diagnosing failures across the complete perception pipeline.
  • Design, implement, and maintain production C++ perception components for camera-based detection, tracking, target localization, and trajectory estimation on autonomous drones.

  • Develop and improve multi-object tracking, data association, motion models, filtering, uncertainty propagation, track management, and fusion of detector and single-object-tracker outputs.

  • Estimate target position and motion in world coordinates by fusing monocular vision with vehicle-state and navigation data, accounting for timing, uncertainty, latency, and degraded sensor states.

  • Integrate and optimize neural-network models in the production C++ pipeline using ONNX, TensorRT, CUDA, and NVIDIA Jetson.

  • Analyze recorded flight data and systematically reproduce and diagnose field failures.

  • Define robust software interfaces and engineering standards, conduct design and code reviews, communicate clearly across disciplines, and collaborate effectively with ML, autonomy, simulation, and flight-test teams.
  • Strong programming skills and experience designing, testing, and maintaining performance-sensitive production software.

  • Hands-on experience shipping perception, state-estimation, localization, or tracking software on real robotic systems.

  • Strong understanding of linear algebra, rigid-body transformations, coordinate frames, and projective geometry.

  • Experience with probabilistic estimation, Kalman filtering or related estimators, uncertainty modeling, motion models, and sensor fusion.

  • Ability to integrate learned models into a broader perception system and reason about preprocessing, output semantics, failure modes, latency, and model-to-system trade-offs.

  • Good understanding of algorithms and data structures, strong software architecture judgment, and a track record of independently owning complex technical systems.

  • Strong problem-solving and communication skills, with the ability to mentor engineers and collaborate across perception, ML, autonomy, embedded, and flight-test teams.
  • Experience with aerial / drone (UAV) perception, target tracking, geolocation, or autonomous aircraft.

  • Familiarity with ROS 2, PX4, ArduPilot, MAVLink, or comparable robotics and flight-control software stacks.

  • Edge / real-time deployment experience with NVIDIA Jetson, TensorRT, CUDA, VPI, GStreamer, or hardware-accelerated camera pipelines.

  • Visual-inertial odometry (VIO), visual SLAM, structure-from-motion, factor graphs, bundle adjustment, or GNSS-denied navigation.

  • Experience with NVIDIA Isaac Sim, software-in-the-loop / hardware-in-the-loop testing, synthetic data, or simulation-based validation.

  • Experience with low-light, thermal, EO/IR, fisheye, gimbal-mounted, or long-range imaging systems.
Real impact & ownership: Shape high-tech defense systems and take responsibility from day one

Mission-driven environment: Work on technologies that matter for European security and sovereignty

Competitive package: Competitive salary, EGYM Wellpass, corporate benefits, and equity options aligned with role and level

Deep tech environment: Get hands-on experience with cutting-edge drone technology and real operational use cases

Grow fast: Steep learning curve for juniors, strategic influence and leadership opportunities for seniors

Startup mindset meets defence innovation: flat hierarchies, fast decisions, and space for your ideas.

Flexibility: Flexible hours, remote options, and relocation support

Strong team: International, driven colleagues and a culture built on exchange, trust, and shared success