Song Chang-hyun "We Will Completely Change the Car Development Method"... The Secret Behind Hyundai Motor Group's Accelerated Shift to SDV

Keynote Speech at HMG Developer Conference on 13th
Developers from Various Fields of Hyundai Motor Group as Speakers
4 Areas: SDV, AI, Autonomous Driving, Mobility Services
Sharing Software Development Ideas and Current Status

Song Chang-hyun, Head of the SDV (Software Defined Vehicle) Division (President) at Hyundai Motor Group and CEO of Forty2Dot, who is leading the group's transition to SDV, has declared a major shift in vehicle development methods. The plan is to move away from the traditional approach where software (SW) is dependent on hardware (HW), and instead separate each development process while simultaneously advancing modularization and standardization to accelerate development speed.

On the 13th, Song announced this plan during his keynote speech at the HMG Developer Conference held at COEX in Seoul. In the automotive industry, SDV is often defined as a "smartphone on wheels" or an "extension of the smartphone ecosystem." However, Song defines SDV as a "fundamental change (paradigm shift)" in the way cars are developed.

Song Chang-hyun, Head of the SDV Division at Hyundai Motor Group and CEO of FortyTwoDot, is delivering the keynote speech at the HMG Developer Conference held at COEX on the 13th.
[Photo by FortyTwoDot]

Song stated, "Currently, all automakers aim to transition to SDV, but they have not found answers in the execution methods," adding, "This is because vehicle development is centered on HW, and there is a lack of understanding of SW development methods." He emphasized that SW development methods must be directly applied to vehicle development. This shift in development methodology itself is what SDV means. To this end, Song summarized the future direction of development changes within the group into four key points.

First, the existing HW-centric vehicle controller structure will be transformed into a structure separated into HW and SW. Previously, the vehicle body was developed first, and then SW was layered on top. Now, from the development stage, HW and SW will be divided and managed as separate projects. This allows for the internalization of each technology or collaboration to be chosen and executed from a business perspective. Second, commonly used SW logic in development will be modularized into a single unit. Creating essential SW logic as one block makes it easier to reuse and shortens verification time.

The third point is standardization. By standardizing communication protocols and data formats between vehicle SW and HW, developers establish a kind of agreement. Standardization also enables simultaneous development and verification of numerous SW components. The final method is the active utilization of open source. Using externally verified open source SW can speed up development and enable connection and collaboration with new external ecosystems.

Song said, "We want people to focus on more valuable tasks rather than being tied to tools of mobility," adding, "SW and AI will inject infinite possibilities into the mobility industry, and this is the responsibility of developers."

Why Has Hyundai’s Genuine Navigation Improved Recently?

At the event, developers from Hyundai Motor Group also took the stage to introduce various vehicle SW development statuses. Sang-kyu Son, Head of the AI Technology Team at Hyundai AutoEver, explained the AI technology used in Hyundai and Kia’s genuine navigation systems.

Until a few years ago, the navigation systems installed in Hyundai and Kia vehicles were criticized for poor accuracy in estimated arrival times. However, recently, Hyundai and Kia owners have been praising the significant improvements in the completeness of the genuine navigation system. The reason behind this positive response is AI technology. Son explained, "To derive an error-free estimated arrival time, complex and diverse variables must be considered," adding, "We utilized artificial intelligence deep learning technology to leverage complex patterns based on data."

Calculating route-based estimated arrival times in navigation requires complex computations. Not only must errors be minimized, but inference speed must also be at a level that allows real-time updates. Son said, "By applying the AI model, the number of large errors with differences exceeding 10 minutes decreased by 74% compared to before," and "Accuracy improved across all time periods."

What If Someone Attacks My Car? ... The Growing Importance of Cybersecurity

Vehicles operating based on SW also face critical cybersecurity challenges. Until now, cars were not attractive targets for cyber attackers compared to financial institution servers or IT devices. However, with the advent of the connected car era, cybersecurity for vehicles has become more important than ever.

Ji-yong Han, Head of Vehicle Cybersecurity Development Team at Hyundai Motor and Kia, said, "Attacks that are outdated in the IT industry can still be new attacks in the automotive field," adding, "Our task is to identify potential anticipated attacks and develop solutions for them."

Europe has already implemented strict regulations since last year, banning the sale of vehicles without cybersecurity certification within the European market. Hyundai and Kia have released vehicles that meet Europe’s security standards and have recently expanded their cybersecurity personnel and organizations.

Han explained, "We have established a response system that detects abnormal attacks on the network, sends them back to the server to create solutions, and then sends these solutions back to the vehicle," adding, "This system has been installed in vehicles sold in Europe as well as new models like the GV60 and Ioniq 6 to strengthen cybersecurity capabilities."

This content was produced with the assistance of AI translation services.