算法研发工程师 Algorithm R&D Engineer (SLAM / Computer Vision)

Company: Universal Tech Limited

Location: Hong Kong Science Park (Sha Tin), N.T. | Shenzhen Commuter Friendly: 1-hour direct commute via East Rail Line to University Station + Bus 272K.

Employment Type: Full-time

Salary Range: HK$16,000 – HK$23,000 / month

Application Deadline: 31 August 2026

Contact: [email protected]

Languages: Mandarin / English

About Us

Universal Tech Limited is an innovative technology company incubated at Hong Kong Science Park, specializing in the core R&D of algorithms for home cleaning robots. Adopting a cross-border R&D model between Shenzhen and Hong Kong, we offer a flexible working environment for GBA talents and provide comprehensive support for Hong Kong work visa renewals (IANG / TTPS / QMAS).

Research & Development Focus

This position belongs to our core R&D initiative: "Full-scene Navigation and Environmental Perception for Home Cleaning Robots." You will focus on solving industry challenges such as localization drift, map relocalization, and obstacle avoidance in complex household environments (e.g., low-light conditions, cluttered layouts, dynamic obstacles).

Key Responsibilities (R&D Oriented)

1. SLAM Algorithm Development (Laser / Visual Navigation)

• Research and develop SLAM algorithms (LiDAR, Visual, and sensor fusion) for home cleaning robots, covering core modules: mapping, localization, loop closure, path planning, auto-docking, and resuming cleaning tasks.
• Enhance algorithm robustness against household-specific challenges to mitigate issues like map ghosting and pose estimation loss.
• Port and optimize algorithms on low-power embedded platforms (e.g., Allwinner, Rockchip, Jetson Nano), balancing localization accuracy with CPU/memory constraints.

2. Environmental Perception & Vision Processing

• Develop algorithms for obstacle detection and recognition using RGB-D cameras, ToF, structured light, and ultrasonic sensors in complex household environments.
• Construct diverse household datasets (covering various lighting, layouts, and floor materials); design and execute Benchmark tests to validate perception capabilities under extreme conditions (low-light, backlight, fast motion).
• Implement multi-sensor fusion algorithms (IMU, wheel odometry, visual, LiDAR) to improve localization stability during slippage, climbing thresholds, or external collisions.

3. Mass Production & Technical Documentation

• Collaborate with hardware teams on sensor calibration and consistency tuning to address mass production variances.
• Draft high-quality technical documentation, test reports, and patent applications to secure intellectual property rights.
• Support system integration and production line testing; rapidly troubleshoot and resolve algorithm issues during pilot runs and mass production phases.

Requirements

Qualifications & Compliance

• Degree: Master's degree or above in Computer Science, Automation, Robotics, Applied Mathematics, Electronic Engineering, or other STEM-related disciplines. (Bachelor's degree holders must have at least 1 year of relevant R&D experience)

• Work Rights: Must possess legal right to work in Hong Kong, including:

  • Hong Kong Permanent Resident
  • IANG (Immigration Arrangements for Non-local Graduates)
  • TTPS (Top Talent Pass Scheme)
  • QMAS (Quality Migrant Admission Scheme)

Hard Skills

• Programming: Proficient in C++ and Python with solid knowledge of data structures and algorithms; experienced in Linux development.
• SLAM Expertise: Deep understanding of mainstream SLAM frameworks (e.g., Cartographer, LOAM, ORB-SLAM, VINS-Fusion); familiar with LiDAR/Visual Odometry, loop closure detection, and Bundle Adjustment optimization.
• Tools: Skilled in PCL, OpenCV, Eigen, Sophus; proficient in ROS/ROS2 (node development, Bag debugging); mastery of Git and CMake.
• Embedded Awareness: Basic understanding of ARM architecture and embedded Linux workflows; strong awareness of performance optimization.

Preferred Experience

• Proven R&D experience in home cleaning robots or service robots with mass production exposure.
• Familiarity with deep learning for perception (YOLO, Transformer) and deployment frameworks (TensorRT, NCNN).
• Proficiency in ARM NEON, CUDA optimization, or algorithm acceleration on low-power platforms.
• Publications in top-tier conferences (CVPR, ICCV, ECCV, ICRA, IROS) is a strong plus.
• Excellent English technical writing and communication skills.

What We Offer

• Visa Sponsorship: Full support for Hong Kong work visa extensions (IANG / TTPS / QMAS).
• Strategic Location: 1-hour direct commute from Luohu Port, ideal for talents residing in Shenzhen.
• Product Impact: Work on core algorithms for consumer products with global market reach.
• R&D Ecosystem: Access to state-of-the-art laboratories and resources at HK Science Park.
• Compensation: Competitive salary package (HK$16,000–23,000/month) with Hong Kong's favorable tax regime.

How to Apply

Please send the following documents to [email protected]:

1. CV/Resume
2. Supporting Materials: Academic papers, project reports, GitHub repositories, or demo videos.

Email Subject: Application for Algorithm R&D Engineer (SLAM / Computer Vision) – [Your Name]