Lead - Perception (UAVs)

Unmannd; Bengaluru; On-site; INR 500,000 - 800,000; Full time

Benefits offered by this job

Access to UAV hardware and flight testing

Significant growth opportunities

High-performance compute resources

Job summary

A leading UAV technology company is seeking a Perception Lead to own the perception architecture for its UAVs. This role involves the development of high-performance perception systems, requiring deep technical expertise and leadership to guide teams. Candidates should have 6–10 years of experience in perception for robotics or drones, with strong skills in sensor fusion and embedded systems. The company offers growth opportunities in a fast-paced startup environment in Bengaluru.

Qualifications

• 6–10 years of experience in perception for robotics, drones, or autonomous vehicles.
• Proven expertise in SLAM and mapping frameworks such as ORB-SLAM and Cartographer.
• Hands-on experience with embedded systems and GPU acceleration.
• Experience with SLAM and mapping frameworks (ORB-SLAM, Cartographer, RTAB-Map, LOAM).
• Hands-on experience with embedded systems & GPU acceleration (CUDA, TensorRT, Jetson, etc.).
• Experience in C++ and Python for robotics software development.
• Track record of deploying perception algorithms on real robots/UAVs.

Responsibilities

• Lead development of UAV perception pipelines for object detection, tracking, mapping, and obstacle avoidance.
• Architect and optimize multi-sensor fusion frameworks.
• Develop algorithms for SLAM, 3D reconstruction, and terrain awareness.
• Ensure perception stack runs in real-time on embedded hardware.
• Develop and deploy ML/DL models for vision-based detection and scene understanding.
• Implement robust algorithms for adverse environments.
• Optimize latency, robustness, and power efficiency in embedded systems.
• Collaborate with autonomy and controls engineers to integrate perception outputs.
• Define perception testing protocols in simulation and real-world tests.
• Analyze UAV flight logs to improve perception performance.
• Stay at cutting edge of computer vision, robotics perception, and AI for autonomy.
• Evaluate and integrate state-of-the-art techniques for certifiable safety.
• Contribute to long-term strategy for safety-critical perception.

Skills

Sensor fusion

SLAM and mapping frameworks

Embedded systems & GPU acceleration

C++

Python

C++ & Python

Robotics software development

Education

Master’s or Ph.D. in Robotics, Computer Vision, Machine Learning, or related field

Tools

CUDA

TensorRT

Jetson

ORB-SLAM

Cartographer

RTAB-Map

LOAM

Job description

We're building deeply intelligent UAVs for defense forces - designed to out-think, outmaneuver, and dominate even in contested environments. Our mission is to give the armed forces an undeniable edge and help safeguard peace across borders. We move fast, challenge limits, and obsess over building technology that redefines what’s possible. If you are driven to build something truly transformational — then this is it.

Job Description

About the Role

The Perception Lead will own the perception architecture for Unmannd UAVs, leading development across vision, LiDAR, radar, and sensor fusion pipelines. You will lead the development of high-performance perception systems that integrate seamlessly with autonomy and control. This role requires both deep technical expertise and leadership skills to guide a cross-functional team.

Key Responsibilities

• Lead development of UAV perception pipelines for object detection, tracking, mapping, and obstacle avoidance.
• Architect and optimize multi-sensor fusion frameworks (e.g., combining cameras, LiDAR, radar, GPS, IMU).
• Develop algorithms for SLAM, 3D reconstruction, and terrain awareness.
• Ensure perception stack runs in real-time on embedded hardware.
• Develop and deploy ML/DL models for vision-based detection, semantic segmentation, and scene understanding.
• Implement robust algorithms for adverse environments (low-light, rain, fog, GPS-denied zones).
• Optimize for latency, robustness, and power efficiency in embedded systems.
• Work with autonomy engineers to feed perception outputs into planning and navigation modules.
• Collaborate with controls engineers to enable perception-informed control loops (e.g., obstacle avoidance).
• Define perception testing protocols in simulation (Gazebo, AirSim, Isaac Sim) and real-world flight tests.
• Analyze UAV flight logs to improve perception performance and robustness.
• Stay at the cutting edge of computer vision, robotics perception, and AI for autonomy.
• Evaluate and integrate state-of-the-art techniques (e.g., deep learning for perception, graph-SLAM, event-based vision).
• Contribute to long-term strategy for making UAV perception certifiable and safety-critical.

Requirements

Qualifications

Master’s or Ph.D. in Robotics, Computer Vision, Machine Learning, or related field.

6–10 years of experience in perception for robotics, drones, or autonomous vehicles.

Strong expertise in:

• Sensor fusion (EKF, UKF, particle filters).
• SLAM and mapping frameworks (ORB-SLAM, Cartographer, RTAB-Map, LOAM).
• Hands-on experience with embedded systems & GPU acceleration (CUDA, TensorRT, Jetson, etc.).
• Experience in C++ and Python for robotics software development.
• Track record of deploying perception algorithms on real robots/UAVs.

What We Offer

Ownership of Unmannd’s perception architecture and roadmap.

Direct impact on enabling safe autonomous flight in real-world environments.

Access to UAV hardware, flight testing, and high-performance compute resources.

A startup environment with significant growth opportunities.