Jiancai Hou – Robotics & Embedded Systems Portfolio

Robotics & Course Projects

Unitree Go2 — Reinforcement Learning Navigation

Dual-pipeline RL (Gazebo TD3 + Isaac Lab PPO), obstacle avoidance & manuscript in preparation

Accepted for oral presentation (ICECC 2026) Springer LNEE Proceedings EI Compendex & Scopus indexed Unitree Go2 RL Navigation Deployment & Edge Offloading

Built a dual-pipeline RL framework for Go2: CNN-based TD3 in Gazebo and hierarchical PPO in Isaac Lab.
Focused on obstacle-avoiding navigation with LiDAR down-sampling, reward shaping, and curriculum design.
Evaluated deployment options: on-robot inference, Ethernet tethering, and 5G edge offloading.
Preparing a research article based on these experiments.

Simulation-Based Development for a Nonlinear Excavator System

Thesis Project · simulation-first workflow for model-based and learning-based control

The thesis is expected to be published in May 2026 Nonlinear System CasADi RL Simulator MPC Simulator

Built a simulation-first development workflow for a nonlinear excavator to support model-based controllers and learning-based controllers.
Formulated MPC using CasADi and nonlinear optimization; developed a motion trajectory task for evaluation.
Prepared an RL-ready control pipeline for motion and trajectory tasks at the environment level; training,experiments.

Open Duck Mini — RL Gait Runtime (Course Project)

Raspberry Pi Zero 2 W · 14 motors · MuJoCo training & ONNX deployment

Quadruped Reinforcement Learning ONNX Runtime Python

Open Duck robot is a replication project with 14 motors and IMU sensors and driving by PPO reinforcement learning model, I am responsible for these sensors’ calibration and processing.
Built a small open-source walking robot runtime using a Raspberry Pi Zero 2 W and 14 motors.
Separated runtime into policy execution, state estimation, and low-level motor control for easier iteration.
Open-source repo: Open Duck Mini Runtime

Internships in Finland — Digital Twin, Care Robot & Go2 Research

Savonia UAS · Thinking Lifeline · University of Turku

ESP8266 / ESP32 Azure · Unity3D Intel RealSense RPlidar A1 ORB-SLAM3 / RTAB-Map ROS2 Navigation2 YOLOv8 Unitree Go2 RL Navigation

Embedded/IoT Engineer Intern @ Savonia UAS (Jan–May 2024) — contributed to a hospital digital twin project by developing a small simulator and integrating ESP8266/ESP32 sensors, Azure cloud services, and Unity3D visualization for patient flow and equipment status.
Embedded Software Engineer Intern @ Thinking Lifeline (May–Sep 2024) — developed a care robot (iRobot-based) with RealSense + RPlidar; integrated ORB-SLAM3/RTAB-Map with ROS2 Navigation2; implemented YOLOv8-based human distance detection and experimented with sound-based navigation.
Robotics Developer Intern @ University of Turku (May–Aug 2025) — trained a dual RL navigation pipeline for Unitree Go2, focusing on perception-driven obstacle-avoiding navigation; manuscript in preparation.

IoT Room Air Temperature & Humidity Monitor

ESP8266 sensors · Cloud DB · Docker · React dashboard

IoT ESP8266 Docker React

Designed an end-to-end IoT system with ESP8266 + environmental sensors sending data to a cloud database.
Containerised backend services with Docker and visualised data in a React-based dashboard.

Custom Mobile Robot with GPS, LiDAR & Odometry Fusion

Course project — sensor fusion for outdoor navigation

GPS LiDAR Odometry Sensor Fusion

Built a practical robot project integrating odometry data, and GPSdata for outdoor navigation.
Implemented data fusion and tested localisation during outdoor motion experiments.

Experience

C++ / Qt Engineer — Xi’an, China

2017 – 2022 · Intelligent Security · Satellite Signals · Radar & Traffic

ChinaSoft International (Mar 2017 – Sep 2019) — Large-scale C++/Qt desktop application for intelligent security monitoring (face & vehicle analysis). Focus: vehicle analytics module, API/data integration, protocol analysis, and feature migration from legacy web to Qt client. Practiced structured engineering workflows: code reviews, task tracking, and cross-team delivery.

Rayco Electronic Information Technology (Sep 2019 – Oct 2020) — Windows C++ applications for transmitting/receiving satellite signals generated by FPGA-based simulators; database storage and real-time visualization with high throughput requirements.

Xi’an Hengsheng Anxin (Nov 2020 – Sep 2022) — Urban traffic monitoring system combining vehicle traffic flow data with radar data; edge deployment on NVIDIA Jetson; TCP data exchange protocol, stream processing, Qt visualization UI, and device calibration via graphical vehicle simulation scenes.

Demo video:

▶ Open on YouTube

Robotics & Embedded Internships — Finland

2024 – 2025 · Savonia UAS · Thinking Lifeline · University of Turku

Moved into robotics and embedded systems through three internships covering hospital digital twins, care-robot navigation, and quadruped research. Work included ESP32/ESP8266 sensing, Azure & Unity3D visualization, RealSense/RPlidar-based SLAM + ROS2 Navigation2, YOLOv8 perception, and RL navigation for Unitree Go2.

Certifications

Professional Certificates

Cloud · DevOps · Project Management

Docker Foundations Professional Certificate (LinkedIn Learning & Docker)

Microsoft Azure AI Essentials Professional Certificate (Microsoft & LinkedIn)

Foundation Certificate in Project Management (PRY · Finland)

Activities

Nokia · Elisa · Microsoft Kuopio Student Network Challenge 2024

Savonia University of Applied Sciences

Participated in the Kuopio Student Network Challenge 2024 organised by Nokia, Elisa and Microsoft, gaining experience in industry-oriented teamwork and presenting technical ideas.

Nokia · Elisa · Microsoft Kuopio Student Network Challenge 2024 at Savonia UAS