Presenting Practical Application Standards for SDV, Edge AI, and Physical AI

Technology Conference Answering Questions from Industrial Field Receives High Praise

"More important than knowing technology is making decisions about where to apply it"

The '2026 e4ds Tech Day - SDV & Edge AI Conference' held on the 8th at Conference Rooms 1 and 2 of the Korea Science and Technology Hall in Seoul concluded successfully with industry representatives, research institutions, and developers gathering in one place.

Going beyond simple technology introduction, this event focused on discussing how Software-Defined Vehicles (SDV), edge AI, and Physical AI should be applied in actual industrial settings in terms of what standards and directions to follow.

Jointly hosted by e4ds news and Channel5 Korea, this event continued with keynote speeches centered on SDV and edge AI, followed by technology contest awards and winning project presentations, surveying the present and future of the technology ecosystem.

Myung Se-hwan, CEO of e4ds, who opened the event, presented the direction of the conference through his opening remarks, stating "Rather than announcing new technologies, we want to address questions that remain unresolved in the field together."

He said, "We focused on presenting practical design standards for questions such as whether it is truly efficient to run AI on an MCU, whether SDV remains a valid strategy, and whether all data must be connected to the cloud to have meaning."

He also presented three keywords to participants—"decision-making criteria, people, and the next step"—urging them to bring home insights that are actually applicable in industrial settings.

The first keynote speech was delivered by Gwak Su-jin, Director of the AIDV Platform Research Division at the Korea Automobile Research Institute, presenting the direction of the future automobile industry on the topic of 'Automotive Software Platform Evolving with AIDV'.

Director Gwak identified 'Physical AI' as a core keyword penetrating the recent global industry, explaining that AI is evolving beyond simple recognition and analysis stages to a stage where it performs judgment and action in the actual physical world.

He cited autonomous vehicles, robots, and smart factories as representative examples of this change, diagnosing that a new industrial ecosystem combining sensors, actuators, and AI software is rapidly forming.

In particular, he analyzed that global competition centered on AI, robots, and autonomous driving is beginning in earnest, with the United States and China engaged in fierce competition to secure leadership in Physical AI.

He also introduced major CES 2026 trends and global autonomous driving industry trends, forecasting that SDV and AI-based mobility platforms will determine core competitiveness in the future automotive industry.

He anticipated that cooperation between automakers and big tech companies will expand further with the advancement of autonomous driving technology.

In the second keynote speech, Shin Dong-joo, CEO of Mobilint, made a presentation on the topic of 'Physical AI Era: AI Semiconductor Technology and Market Trends'.

CEO Shin explained that the evolution of AI technology is entering the Physical AI era through Perceptual AI centered on recognition and generative AI, followed by Agentic AI focused on autonomous execution.

He emphasized that the core of AI competitiveness going forward lies not in developing larger models but in inference-focused technology that actually operates in real-time in the field.

He particularly compared cloud-centric AI with on-device AI, explaining that edge AI semiconductors with ultra-low latency, autonomous operation, and security are essential for Physical AI to be implemented.

For drones, robots, autonomous vehicles, and industrial equipment to perceive and make decisions on their own, NPU-based AI SoC technology capable of immediate inference at the site is important.

He also forecasted that the future AI semiconductor market will shift its center of gravity from learning-focused to inference-focused, and that edge AI platforms with both energy efficiency and versatility will drive changes across industries.

The final session of the morning was marked by the 'TechDay Contest 2026' awards ceremony.

This contest was established to discover projects that have actually solved problems using edge AI technology, and the organizers expressed their intention to create an ecosystem where technology accumulates as verifiable results rather than mere information.

The awards in this ceremony were presented respectively by Mouser Electronics and STMicroelectronics, both platinum sponsors of e4ds Tech Day 2026.

Daphne Tien, Vice President of Mouser, and Matteo Marabita, Center Director of STMicroelectronics APAC AI Competence Center, presented awards directly to the winners.

Winners received prize money of 1 million won each.

The winning projects selected were LightGuard by Engineer Park Min-hyun and an AI intrusion detection system based on EBC7700 by Engineer Bang Jeong-bae.

Winners Engineer Park Min-hyun and Engineer Bang Jeong-bae shared their development processes and application cases through project presentations.

The winning projects, which addressed safety and security issues using edge AI in actual field settings, received significant attention from participants.

Related projects can be found on the e4ds Makers platform.

The e4ds Tech Day 2026 was an occasion that presented what the industry should prepare for amid the massive technological changes of SDV, edge AI, and Physical AI.

Above all, it held significance in that it focused on the question of 'where, why, and how to apply' rather than on technology itself.

e4ds plans to move forward beyond conveying technical information to connecting collective intelligence from the industrial field through actual projects, contests, and technology verification platforms, and to build an ecosystem that proves the practical value of new technologies.