Bangkok Metropolitan Area
{RESEARCH - COPY - CUSTOMIZE - IMPROVE} is how I work. - Conducts research and development on applied AI software, specifically focusing on Computer Vision and related systems. - Experienced in Tollway's Realtime License Plate Reading Software Deployment on Resources Constrained System and Vehicles Identification development. [GPA 3.75] 1st honor degree in Robotics and AI Engineering [GPA 3.95] MBA - Understand AI from ground up and able to manipulate architectures, pipelines, and libraries. - I listened and research, then speak facts I gathered, not just work by commands. - Aims to build Engineering team with both Business and Engineer Mindset. - Now adding LLMs to Skills set.
Supported the end-to-end design, development, and testing of robotics and AI solutions to enhance organizational capabilities and deliver results aligned with client business objectives.
Integrating AI algorithms and Computer Vision with Point Cloud technology to improve building health inspection's speed and accuracy for phd program. #pointcloud #computervision #civilengineering #research
As the lead of a four-member software development team, I managed and developed two key solutions for tollway systems utilizing YOLOv10 models and custom-optimized libraries. The first project was License Plate Recognition (LPR) software, specifically fine-tuned for efficient operation on CPU-only servers. We modified the EasyOCR library by decoupling detection models from the algorithms, which significantly reduced processing time and enhanced accuracy. The system achieved over 95% accuracy in reading characters and province names, exceeding customer requirements. The second project was an Automatic Vehicle Classification (AVC) system, prioritizing precise vehicle entry/exit detection and wheel counting to ensure system integrity. The team was responsible for the entire scope, ranging from requirement analysis and hardware planning to a hybrid data labeling process (combining manual and semi-automated methods). The resulting system reduced maintenance overhead, offered high flexibility, and delivered accurate performance in real-world environments.
Lead a team size of 4 develop a video streaming license plate reading software (detect vehicle using object detection instead of sensors) with an accuracy more than 95% both license and province, and research on Automatic Vehicle Classification for toll collection using wheels counting.
During my Industrial Internship in the Expressway Department at Loxley Public Company Limited, I conducted in-depth studies on toll collection technologies and security systems utilized within the expressway industry. This included experience with closed toll systems employing automatic license plate recognition on the Burapha Withi Expressway, as well as the implementation of traffic anomaly detection systems on the M6 and M81 Motorway projects. As part of my Cooperative Education program, I developed a system to monitor, collect data, and generate alerts regarding traffic volumes for outbound toll lanes 1–4 at the Bangna Km.6 toll plaza. Leveraging Computer Vision Deep Learning techniques, the system was capable of accurately detecting and tracking vehicles to calculate both traffic volume and speed, which were then used to trigger traffic condition alerts. To further enhance efficiency, multiprocessing techniques were applied to minimize processing times and reduce output latency.