IVIS Joins Hyundai Mobis-Led AI Vehicle Computer Project

IVIS Joins Hyundai Mobis-Led AI Vehicle Computer Research Project

South Korea-based automotive software company IVIS will participate as a joint research and development institution in a project focused on multimodal sensor fusion and high-precision synchronization technology for high-performance vehicle computer validation. The project forms part of the Ministry of Trade, Industry and Resources' automotive industry technology development program and contributes to the broader AI-Driven High-Performance Computer Development initiative. Led by Hyundai Mobis, the research program is scheduled to continue through December 2029. Its central objective is to advance the computing and sensor-processing capabilities required to support next-generation autonomous driving systems.

AI-Driven High-Performance Vehicle Computer Targets 2,000 TOPS

The broader initiative aims to develop a high-performance vehicle computer, or HPVC, with artificial intelligence computing performance of up to 2,000 trillion operations per second, expressed as 2,000 TOPS. This computing target provides the technical context for the project's work on sensor data processing and synchronization. The research is intended to support autonomous driving based on the Vision-Language-Action, or VLA, approach, which connects visual perception, language-related understanding, and action-oriented decision-making. The source announcement does not specify the final computer's architecture, deployment schedule, or commercial vehicle applications, so these details remain unconfirmed.

Multimodal Sensor Fusion and High-Precision Synchronization

The project will develop technology for processing and synchronizing data in real time from multiple vehicle sensors, including cameras and radar. Combining information from different sensor types requires data to be coordinated accurately so that the vehicle computer can work with inputs representing the driving environment. High-precision synchronization is therefore a central part of the validation effort, alongside multimodal sensor fusion. The project also plans to validate the technology in actual vehicles, providing a practical setting for evaluating its operation beyond software development alone. The announcement does not disclose specific synchronization tolerances, sensor configurations, or validation results.

IVIS to Develop the HPVC Software Environment and Middleware

IVIS will establish the software development environment for the high-performance vehicle computer and take responsibility for applying, validating, and optimizing middleware that connects in-vehicle hardware with software. Middleware provides an integration layer between components, helping different parts of a vehicle's computing system communicate and operate together. Within this project, IVIS's responsibilities focus on the software infrastructure needed to develop and validate the HPVC platform, rather than on a separately announced production vehicle or consumer-facing product. The source announcement does not identify the specific middleware technologies or software interfaces involved.

Project Timeline and Development Scope

The Hyundai Mobis-led project is scheduled to run through December 2029 under South Korea's automotive industry technology development program. IVIS will contribute as a joint R&D institution, with work centered on the HPVC software development environment and middleware application, validation, and optimization. The project's scope combines sensor-data processing and synchronization research with validation in actual vehicles, while the wider initiative targets AI computing performance of up to 2,000 TOPS. The announcement does not provide interim milestones, a budget, named vehicle models, or a date for commercial deployment. Those details should not be assumed from the stated project end date.

Industry Impact & Outlook

The project brings automotive software integration, sensor-data synchronization, and high-performance AI computing together within South Korea's vehicle technology development effort. If the technology is successfully developed and validated, it could help address the software and data-coordination requirements of VLA-based autonomous driving systems. IVIS's role highlights the importance of middleware and development environments alongside computing hardware and sensor technologies, while Hyundai Mobis leads the overall project. The December 2029 timeline establishes the planned research horizon, but the announcement does not confirm when the resulting technology will enter production vehicles or what commercial systems will use it.

Frequently Asked Questions

What role will IVIS play in the HPVC development project?
IVIS will participate as a joint R&D institution, focusing on the software environment and middleware needed for high-performance vehicle computer development. Its responsibilities include building the HPVC software development environment and handling the application, validation, and optimization of middleware connecting in-vehicle hardware and software. These tasks support integration and testing within the broader project. The announcement does not assign IVIS responsibility for leading the entire initiative, developing a specific production vehicle, or delivering a commercially available autonomous driving system.

What is the target AI computing performance of the HPVC?
The broader AI-Driven High-Performance Computer Development initiative aims to develop a high-performance vehicle computer with AI computing performance of up to 2,000 TOPS. TOPS refers to trillions of operations per second and is used to describe computing throughput. This figure is the initiative's stated target, not confirmation that a completed system has already achieved that performance. The announcement does not provide further specifications for the computer, including its processor configuration, power consumption, memory, or final implementation in a vehicle.

What technology will the project develop and validate?
The project will develop multimodal sensor fusion and high-precision synchronization technology for validating a high-performance vehicle computer. It is intended to process and synchronize real-time data from multiple sensors, including cameras and radar, and to validate the technology in actual vehicles. These activities are designed to support autonomous driving based on the Vision-Language-Action approach. The announcement does not disclose detailed technical performance targets, specific sensor models, synchronization tolerances, or completed test results, so the effectiveness of the technology cannot yet be determined from the stated information.

What does VLA-based autonomous driving mean in this project?
VLA stands for Vision-Language-Action and describes the approach to autonomous driving that the project aims to support through improved computing and sensor-data processing. In this context, the work focuses on coordinating real-time inputs from sensors such as cameras and radar and preparing the HPVC software environment for validation. The announcement identifies VLA-based autonomous driving as the intended application, but it does not detail the specific model design, decision-making architecture, or vehicle functions that the resulting technology will support.

Who is leading the project, and which program supports it?
Hyundai Mobis is leading the project, in which IVIS will participate as a joint R&D institution. The project is part of the Ministry of Trade, Industry and Resources' automotive industry technology development program and falls within the broader AI-Driven High-Performance Computer Development initiative. Its focus is the development of sensor fusion and synchronization technology for high-performance vehicle computer validation. The announcement does not identify additional participating institutions, disclose project funding, or provide a complete list of organizations contributing to the wider initiative.

When is the project expected to finish, and when could the technology reach vehicles?
The Hyundai Mobis-led project is scheduled to run through December 2029, establishing the stated timeline for the research and development effort. The planned work includes developing real-time sensor-data processing and synchronization technology, building the HPVC software environment, and validating the technology in actual vehicles. However, the announced end date does not establish when a commercial product will launch or when the technology might appear in production vehicles. No commercial deployment date, production commitment, or vehicle-specific implementation has been disclosed.