Intro

work is love made visible

Intern with Autoliv Electronics Japan (currently known as Veoneer) [March 2016]

The best scientists and engineers is as creative as the best storytellers.

― Steve Jobs

I thrive on embracing change and unknowns, focusing on key priorities while maintaining flexibility. I deeply value both individual creativity and collaborative team spirit. Balancing the big picture with attention to detail, I make decisions that drive holistic and sustainable impact.

Bringing authenticity, ethical integrity, and harmony into the lives of individuals and communities is at the core of my personal and professional journey.

I am currently interested in creating unique value at the intersection of Electronics, High-performance Embedded Computing, and Software Engineering in the software-defined vehicle and intelligent transportation domain. Technology alone is not enough - we need to establish positive public-private partnership across multiple disciplines to deliver human-centric, holistic solutions.

Full Time

Software Engineer - General Motors of Canada (Jun. 2020 - Present)

Software Defined Vehicle and Operating System (Feb. 2021 - Present)

• Led and facilitated the development of a novel, industry-first test framework in C++ to verify functional requirements and behaviors of the AUTOSAR Adaptive Platform middleware on the next-generation high-performance, ARM SoC-based domain controller family for GM’s next-gen Driver Assistance System (ADAS) Super Cruise and Ultra Cruise.

• Collaborated with the Test & Simulation team on development of VIRES simulation scenarios and integration of Applied Intuition vehicle dynamics simulations in the Hardware/Software-in-the-Loop (HIL/SIL) environments.

• Led resolution activities for the networking and camera-based sensing software issues (TCP/IP, C, C++) and managed to consistently root-cause and solve numerous critical problems on time with quality.

• Collaborated with the software architecture team and application teams globally (US, Israel) to ensure the software is built and integrated with efficiency and quality using Docker container-based solutions.

• Designed and implemented an end-to-end, CI/CD Jenkins pipeline to enable cross-platform software compilation (LLVM, QCC, Docker), Unit testing (GoogleTest), on-target testing, and reporting (Python, Streamlit, Reportportal).

• Ramped up quickly and took on core work within one month, followed and advocated for Scrum processes, improved internal documentation, and mentored 15 new hires within 2 years of service.

Vehicle Motion Embedded Controls (Jun. 2020 - Jan. 2021)

• Designed and developed sensor diagnostic algorithms in MATLAB Simulink and C for in-house brake embedded microcontrollers being used for GM’s next-generation electric and autonomous vehicles.

• Led the Hardware and Software-in-the Loop(HIL/SIL)test strategy of the brake diagnostics algorithms using high-fidelity vehicle and brake system simulations and eliminated 80% of the test cases that were originally performed in vehicle.

• Architected an automated testing framework in Python and implemented high-level abstraction APIs to interact with the test environment, significantly reduced test script development time from an average of 1 hour to 10 minutes.

• Led the core development of a Wheel Speed Sensor Emulator using STM32 MCU in C to generate real-time pulses information to inject faulty behaviors into the vehicle testing environment.

• Developed internal tools in Python and dashboard in Power BI to achieve automated controller instrumentation and calibration, data file storage, and manipulation & analysis.

Internships

Software Developer in Test (Co-op) - FLIR Systems (May - Dec. 2019)

• Developed unit, integration and functional tests for new SDK APIs in C++, C, and Python for internal and external transport layer libraries according to the GenICam GenTL standard.

• Automated code coverage analysis and reporting for unit and functional test suites of the SDK.

• Redesigned and automated software installation, example execution, and GUI applications testing, reducing 70% manual work and resulting in more than 50% overall timesaving for release testing.

• Improved build infrastructure with Azure build pipeline, Docker containers, CMake, compiler cache, and distributed compilation, reducing build time by more than 60% with improved build stability.

• Practiced Agile methodologies and collaborated with Firmware, Customer Support, and Product Management teams as a rotational scrum master to facilitate fast-paced issue resolutions.

Active Safety Engineer (Co-op) - Autoliv Electronics Japan (Jan. - Dec. 2016)

• Programmed ECU gateways in C to communicate between vehicle and external sensor’s CAN signals.

• Developed a data framework in Python to support configuration file parsing, file conversion, data geo-tagging, and visualization to handle TB-magnitude of data recordings from ADAS sensors.

• Created a wireless, real-time, plug-and-play ECU diagnostic device with an Interactive touch interface to replace the original diagnostic software, reducing the setup time from 10 minutes to 30 seconds.

On-Campus

Teaching Assistant - UBC Electrical and Computer Engineering (2019-2020)

• Supervised weekly lab sessions of an undergraduate project design course involving computer hardware and software.

• Evaluated 45 students’ work and guide them through problem-solving processes using active learning techniques.

Learning Space Steward - UBC IT - AV Service (2015-2018)

• Provided technical on-site support for instructors and students on the Audio Visual equipment.

• Inspected the Audio-Visual equipment on campus before classes start to ensure the system’s functionality.

• Maintained classroom seating capacities, layout, and tidiness and reported AV equipment and furniture deficiencies.

Publications

An Efficient Random Access Light Field Video Compression Utilizing Diagonal Inter-View Prediction

2019 IEEE International Conference on Image Processing (ICIP) - Taipei

• Proposed an efficient pseudo-sequence based Light Field video compression prediction scheme offering the best trade-off between compression efficiency and frame random access complexity.

• Reduced 20%/60% (average/worst-case) random access complexity during the video decode stage with minimum bit-rate, reducing buffer size of the playback devices for AR and VR applications.

• Developed an end-to-end, flexible, scalable, and fully automated framework that applies our proposed method to generate spatial inter-view dependency profiles used by the HEVC encoder multiview extension (MV-HEVC).

• Automated the compression execution and result visualization across 5 prediction structures, each with 4quantization parameters (QPs) and 2 video datasets.