Perception Software Intern — Humanoid Robotics

As a Perception Software Engineering Intern at Apptronik, you will work alongside senior engineers and researchers to develop perception systems that enable humanoid robots to understand and interact with real-world human environments. You will contribute to components spanning object detection, tracking, segmentation, and multi-sensor perception, supporting robust autonomy in dynamic settings.

This role provides hands-on experience applying modern computer vision and deep learning techniques to real robotic platforms. You will help build and evaluate perception pipelines, integrate data from multiple sensors, and validate performance in simulation and on hardware. Your work will directly support humanoid navigation, manipulation, and human-robot interaction.

Core Responsibilities

• Assist in developing and testing perception pipelines for humanoid robots, including object detection, segmentation, tracking, and pose estimation.
• Implement and evaluate deep learning models for real-time vision and 3D perception tasks.
• Support multi-sensor data processing and fusion using cameras, depth sensors, LiDAR, and IMUs.
• Contribute to data collection, dataset curation, labeling, and performance evaluation workflows.
• Help integrate perception modules with navigation, manipulation, and control systems.
• Profile and debug perception components to improve latency, accuracy, and robustness.
• Validate perception algorithms in simulation environments and on physical robots.
• Collaborate with cross-functional teams and follow best practices in software development and experimentation.

Required Qualifications

• Currently pursuing a BS, MS, or PhD in Computer Science, Robotics, Computer Engineering, or a related field.
• Coursework or project experience in computer vision, robotics, or machine learning.
• Experience with Python and/or C++.
• Familiarity with at least one deep learning framework (e.g., PyTorch, TensorFlow, JAX).
• Basic understanding of computer vision concepts (detection, segmentation, geometry).
• Ability to work effectively in a collaborative engineering environment.

Preferred Skills

• Experience with OpenCV, ROS/ROS2, or robotics simulation tools.
• Exposure to multi-sensor perception or 3D vision.
• Experience training or evaluating neural networks on real or synthetic datasets.
• Familiarity with Git, testing, and Linux-based development.

Outcome Focus
Deliver robust, real-time perception components that integrate into the full autonomy stack and improve robot reliability and situational awareness.