October 14 '2022 Automotive Tech Concert' Viewer Q&A
ISO TC204, Reflecting Key Perspectives on Road Traffic Systems and Overall Road Environment

Precision road maps are expected to be utilized as an important tool for pre-verifying whether the expansion of tangible and intangible infrastructure can be achieved efficiently.

Yoo Jae-jun, Principal Researcher at ETRI, provided answers on October 14 to questions that could not be addressed due to time constraints during the presentation on "Standardization Trends for Precision Road Maps for Autonomous Driving" at the 2022 e4ds Automotive Tech Concert.

■ What are the key matters that need to be reviewed and verified when attempting to standardize precision road maps for autonomous driving?

Above all, it appears necessary to clearly define the requirements of "what precision road maps should provide for autonomous driving?" This is currently being pursued at ISO TC204 and other bodies, and map specifications that can be utilized as precision road maps reflect key perspectives of △road traffic systems △overall road environment and △applied services. In addition, for vehicle driving and control in autonomous driving, it would be beneficial to reflect requirements from a vehicle perspective, but this is an area where automotive manufacturers' intellectual property rights may be involved, and therefore is not easily disclosed, as I understand it. Nevertheless, it is self-evident that vehicle requirements are necessary to effectively support autonomous driving.

■ What is the method to effectively expand and construct infrastructure for commercializing autonomous driving? Additionally, please explain how to utilize precision road maps effectively and beneficially in this regard.

For effective expansion and construction of infrastructure for autonomous driving commercialization, it is necessary to first define what infrastructure must have and what requirements are needed for autonomous driving, and then determine how to efficiently configure these based on the current status of existing infrastructure. In that process, precision road maps can be utilized as an important tool to accurately describe existing road infrastructure and driving environments as they are, and to pre-verify whether the expansion of tangible and intangible infrastructure can be achieved efficiently.

■ When explaining LDM, traffic signals were also mentioned. Is signal information provision opened, and are signals updated in real-time? If latency occurs, what is the tolerance level?

Message specifications for signal information provision have been established, and some domestic autonomous driving pilot zones are applying and utilizing them in pilot programs, as I understand. Since the signal message is generated and transmitted based on signal information conveyed from the signal controller, it can be said to be updated in real-time. However, in the case of latency, there are aspects that vary depending on implementation, so the exact amount of delay is not known, but since signal information requires real-time transmission, it is not expected to be large. The standard also does not explicitly specify the tolerance for signal delay.

■ At the Korean Academy of Engineering's Autonomous Driving Forum, parking spaces were selected as the priority application field rather than ordinary roads. Are such parking spaces included in the road definitions you described?

Currently, detailed information about parking lots or parking spaces is not included in the scope of precision road maps. As ancillary facilities related to roads, an overview of entire parking lots or parking spaces may be included, but detailed information about parking slots within parking spaces and movement within parking spaces is not yet included. I understand that the National Geographic Information Institute's precision road map intends to include this in the future, and in ISO TC204 standards, this is also being discussed as a separate standard.

■ What is the LoD level currently applied to the implemented map? I am curious whether all components are applied at the same level or differentially.

>> LoD indicates the level of data specificity. Currently, precision road maps do not have LoD separately defined. This is because precision road maps require a specific level of detail so they can be utilized in control and other applications. If LoD were to be applied, it is expected to be applied differently depending on the importance of constituent elements.


■ What is the number of functions provided by default in RoadXML and the method for adding and updating functions?

>> The number of functions that can be defined by default and extended may change depending on discussions and versions, so it is recommended to directly check the specification. Along with this, it is also advisable to directly refer to the mechanism explained in the corresponding specification for the method of adding and updating.

■ I would like to know about recent issues and resolution cases related to precision road map standardization.

There are two most recent issues. First is dynamic transmission and updating of portions of precision road maps, and second is ensuring interoperability with other standardization discussion groups such as ISO TC211 spatial information. Since both are important and challenging issues, it would be more accurate to say that efforts for resolution are continuing rather than that the issues have been resolved. Related standards continue to be discussed for dynamic transmission and updating of portions of precision maps, and continuous standardization discussions are being conducted through JWG (Joint Working Group) with the relevant group to ensure interoperability.