Thanakrit Wongkha

VLA Robotic Arm

Case study

FEATURED / ROBOT LEARNING / 2026

From human intent to robot action.

Led the AI research and solution architecture for a language-conditioned robotic arm with scene-graph representation.

Role: AI/VLA Lead
Year: 2026
Status: Research + build

About Thanakrit

Chonburi, Thailand

ROBOTICS / AI / PRODUCT ENGINEERING

I work where robots, AI, and software meet.

I am a Robotics and Automation Engineer who likes owning the whole path from a wired control cabinet to a working interface. I bring up ROS2 systems, integrate sensing and navigation, prototype AI behavior, and build the tools people need to operate the result.

Right now, I am especially interested in VLA manipulation and autonomous reception robots: machines that can understand intent, perceive context, and act usefully in the real world. I also use Codex, Claude Code, and open-source AI harnesses for software development, with MCP tool integration and multi-agent workflows.

Email me ↗ GitHub ↗

Open to robotics, automation, and applied AI roles

Products & Open Source

6 selected builds

PINNED OPEN SOURCE / PRODUCTS / CLIENT WORK

Proof that ships beyond the robot.

Two pinned public repositories and four end-to-end software builds.

github.com/fordft ↗

Thanakrit Wongkha — Résumé

Updated Sep 2026

ROBOTICS & AUTOMATION ENGINEER

Full-stack robotics builder across hardware, autonomy, AI, and operator software.

Download PDF

Robotics and Automation Engineering graduate with hands-on experience in industrial AMRs, PLC programming, ROS2/OpenRMF fleet systems, VLA and computer vision, web/mobile products, and agentic software development using Codex, Claude Code, and open-source AI harnesses.

Chonburi, Thailand +66 95 756 6669 thanakritwongkha@gmail.com github.com/fordft ↗

Experience

Robotics & Automation Intern

Pro En Innovation Co., Ltd. · Four-month internship

Completed Jan 2026

Owned wiring, robot-control software, and web control for a differential-drive AMR designed for payloads up to 200 kg.
Developed a ROS2/OpenRMF simulation and fleet-management architecture for coordinated robot workflows and software-hardware integration.
Independently built ProEn ERP to digitize purchase-request and purchase-order (PR/PO) tracking and centralize workflow status.

Founder & Product Developer

QAG

Pre-launch

Owns product planning, UX/UI, implementation, testing, domain setup, and deployment end to end.

Freelance Developer

Sangaroon Auto Car

Project-based

Co-developed responsive bilingual vehicle discovery across web and a companion mobile application in a two-person team.

Selected projects

MARS — Mobile Autonomous Reception and Guidance System

Original Prototype Software & Electrical Lead · Final-Year Contributor / Technical Advisor

2024–2026

Implemented the complete software stack and all electrical wiring for the Year 3 prototype; teammates handled mechanical design and assembly.
Architected ROS2 Nav2 navigation with LiDAR, IMU, and RealSense; later contributed AI work and technical advice for the Year 4 extension.

VLA Robotic Arm with Scene Graph Representation

AI/VLA Research & Solution Lead · Final-Year Project Member

2026

Evaluated VLM/VLA models, defined the training strategy, and led solution architecture for a language-conditioned robotic arm.
Configured and trained candidate models with Python, PyTorch, Hugging Face LeRobot, and SmolVLA; translated findings into the team plan.

Auto-Rickshaw Detection Benchmark · IDD

Exploratory Computer Vision Research

2025–2026

Prepared India Driving Dataset data and trained YOLOv8, YOLOv10, YOLO11, and YOLO26; compared validation accuracy, training speed, and compute use to evaluate accuracy-efficiency trade-offs.

Core stack

Robotics

ROS2, Nav2, OpenRMF, differential-drive & holonomic AMRs, fleet management, web-based robot control, LiDAR/RPLIDAR, IMU, Intel RealSense, electrical wiring

PLC & Industrial Control

Mitsubishi PLC (GX Works2, GX Works3), Siemens PLC (TIA Portal), Ladder Logic, I/O configuration, wiring, troubleshooting

AI & Vision

Python, PyTorch, OpenCV, Ultralytics YOLO (v8, v10, YOLO11, YOLO26), VLM/VLA, Hugging Face LeRobot, SmolVLA

Embedded & Tools

C/C++, Arduino, ESP32, Jetson Nano, Linux, SolidWorks, FluidSIM

Product & Software

Web UI/UX, mobile applications, workflow systems, Dart/Flutter, Firebase, Git/GitHub

Agentic Development

OpenAI Codex CLI, Claude Code, Pi Coding Agent, Hermes Agent, OpenClaw, open-source/forked AI harnesses, MCP and tool integration, prompt/context engineering, multi-agent workflows

Recognition

WorldSkills Thailand, PLC Skill — National 3rd Runner-up
1st Place, 6th International Convention on Vocational Students' Innovation Projects (2021)
Bronze Medal, MCOT-ABU National Vocational Robotics Competition (2021)
Semifinalist, Thailand Open Robotic Competition @Home Education (2024)
Participant, IDC Robocon International and Robo Innovator Challenge (2023)

Education

Kasetsart University, Sriracha Campus B.Eng. Robotics and Automation Systems Engineering (International)
2022–2026 · Graduated 2026
GPA: 2.87/4.00

Languages

Thai (Native)
English (Working Proficiency)

I build robots thatsee, move, and act.

I build robots that
see, move, and act.