
On-site Technical Demo
NVIDIA Omniverse / Isaac Sim · Industrial Robotics · Customer Deployment
I build and deploy industrial Physical AI and Digital Twin systems from 0-to-1 development through real-world customer validation, robotics integration, and factory adoption.
Hands-on engineering with customer-facing ownership across Physical AI, robotics and manufacturing.
Primary owner and developer of the first-generation industrial Digital Twin / Physical AI platform at Pegatron, covering NVIDIA Omniverse and Isaac Sim, local applications, third-party integration, robot integration, deployment and cross-system debugging.
Served as the Digital Twin DRI for a major U.S. consumer electronics manufacturing program, translating factory requirements into technical solutions and owning integration, feature delivery, debugging, validation and technical enablement.
M.S. from the Graduate Degree Program of Artificial Intelligence at National Yang Ming Chiao Tung University (NYCU), with graduate research in synthetic data generation using Conditional Normalizing Flows.
A deployment-oriented stack spanning simulation, robots, software and factory integration.
NVIDIA Omniverse, Isaac Sim, OpenUSD, PhysX, Digital Twins, Real-to-Sim, Sim-to-Real, Synthetic Data
ROS 2, MoveIt, Universal Robots / UR5e, Epson, Flexiv, Manipulation, Motion Planning, Pick-and-Place
RTDE, MQTT, AWS IoT Greengrass, REST APIs, industrial equipment and camera integration
Python, C/C++, JavaScript, Linux, Docker, APIs, application UI and deployment tooling
Requirements discovery, solution design, onsite debugging, customer validation, enablement and knowledge transfer
PyTorch, TensorFlow, OpenCV, NumPy, Pandas, Normalizing Flows and synthetic-data research
0-to-1 engineering, customer delivery, productization and factory adoption.
Build and deploy industrial Digital Twin and robotics solutions using NVIDIA Omniverse / Isaac Sim, serving across hands-on development, solution integration, customer delivery and production deployment.
Owned the first-generation Physical AI / Digital Twin platform across simulation, UI, third-party software, robot integration and deployment, before later development was expanded to additional team members.
Owned Digital Twin delivery across four production stations—Pick-and-Place, Screw, Press and robotic dispensing—working directly with customer engineers on requirements, technical decisions, debugging and validation.
Acted as the primary Digital Twin technical owner during 12 total weeks of U.S. onsite support, resolving cross-team integration issues and delivering new features required for acceptance.
Successful customer validation after onsite engagements enabled major commercial milestones, connecting engineering delivery directly with project execution and business outcomes.
Established deployment workflows and transferred technical knowledge so customer engineers could independently extend Digital Twin deployment beyond the initial four stations.
Supported evolution of the original solution toward a reusable product platform and currently contributes to Digital Twin optimization PoCs across two internal Pegatron production lines.
Representing the project from technical demonstration to commercial productization.
March 2025 • San Jose, CA

PEGAVERSE JEDI
Technical DisplayDemonstrated the first-generation system in person, explaining the engineering architecture and industrial use cases to external technical audiences. The project later appeared in NVIDIA public customer and keynote content.
May 2025 • Taipei, Taiwan

Represented the platform as it evolved from an exhibition prototype into a reusable customer solution, with follow-on engagement from the U.S. program and growing interest from manufacturing organizations in Taiwan.
Professional Physical AI systems plus academic and open-source work.
Industrial platform combining NVIDIA Omniverse / Isaac Sim, real-to-sim synchronization, robot integration, machine interfaces and application workflows. I owned the first-generation implementation and later supported its transition into a reusable product and broader factory deployment.
Digital Twin DRI across four production stations. Two six-week onsite engagements covered requirements, solution integration, new feature delivery, debugging, validation and customer enablement. Customer identity and implementation details are intentionally omitted for confidentiality.
Supporting group-level factory teams in evaluating where Digital Twin and simulation can improve line efficiency, process visibility and engineering decision-making. The current phase is PoC and use-case discovery across two production lines.
Graduate research on Conditional Normalizing Flows for high-quality synthetic data generation, including data upsampling, outlier-focused generation and objective evaluation of generated data quality.
Python-based stock analysis and strategy backtesting projects covering data processing, technical indicators, strategy evaluation and performance analysis.
Autonomous patrol vehicle integrating ROS, computer vision, remote monitoring and a web control interface—an early robotics project preceding my current industrial robotics work.
Open to conversations around Physical AI, robotics, simulation and industrial deployment.
For recruiting, technical collaboration or engineering discussions, email or connect with me through LinkedIn.