Inavi Systems shared various information, including ongoing research projects, saying that it is researching and developing technologies required for the era of autonomous driving and the construction of precision maps in the autonomous driving field, which is the core of the future map market.
On the 21st, e4ds hosted a webinar at the 2022 e4ds Automotive Tech Concert, inviting Director Jin-Geun Park of iNavi Systems to present a webinar on the topic of 'Hybrid Map for Point-to-Point Movement Based on Designated Areas'.
iNavi Systems, widely known for its ‘iNavi’ navigation service, has established itself as a ‘map platform’ specialized company that supplies maps to various mobility industries such as taxis, logistics, and delivery.
Director Park said, “In order to provide platform services using the material called ‘map,’ various technologies are needed,” and “iNavi Systems possesses core engines such as field surveys, data construction, and route search, as well as service development technologies.”
Currently, the company is self-driving In relation to this, we are carrying out two autonomous driving government projects led by the Ministry of Trade, Industry and Energy, and are working with several autonomous driving players developing autonomous driving systems (ADS) to provide maps and cooperate on autonomous driving technology.
The presentation topic, designated area-based P2P autonomous driving, is part of the 2021 Ministry of Trade, Industry and Energy government project, 'Designated Area-Based P2P Mobility Lv4 Passenger Vehicle-Class Autonomous Driving Vehicle Platform Technology Development', and iNavi Systems is researching technology to create routes for autonomous vehicles using map building and path-finding technologies and to create trajectory information followed by autonomous vehicles.
P2P autonomous driving is a concept that expands the application of technology that generates autonomous vehicle paths and trajectories using navigation path search technology, and is considered a technology necessary to ensure the free movement of autonomous vehicles.
As part of this technology, iNavi Systems is researching and developing hybrid maps.
▲ Inavi Systems Hybrid Map Configuration The hybrid map consists of the following: △SD Map, a route search map optimized for P2P route generation; △HD Map for generating precise trajectory information; and △Safety Map, which is additional information on the surrounding conditions (such as GPS sound coverage areas) required for safe autonomous driving.
To create a trajectory Since SD and HD Map must have connectivity, Node2Node with directionality of SD Map and link grouping information of HD Map are configured as a single connection information.
Since the technology that iNavi Systems is currently researching and developing is the generation of autonomous driving routes for passenger vehicles, it is necessary to consider the specifications of passenger vehicles and design a safe route and an algorithm to avoid roads where autonomous vehicles cannot drive.
For the above reasons, sharp turning radii, narrow U-turn roads, back roads, destinations immediately after left turns, and child protection zones are designed to be avoided.
Director Park said, “The success factor of autonomous driving is relieving people’s anxiety,” and “Inavi Systems is considering as many avoidable factors as possible during the development stage to relieve anxiety.”
He added that if Level 4 autonomous driving is commercialized in 2027, it should ultimately be advanced to provide the same route as a human driver.
Trajectory refers to the trajectory information that an autonomous vehicle follows.
▲ Trajectory information It generates a trajectory that matches the P2P route by utilizing lane information and driving route line information from the HD Map, and the trajectory is Point Arr.It includes driving limit speed, left and right lane information, and target lane information in the form of ay.
The first key point of Trajectory is that it consists of the simplest Point in map representation and is composed of various attribute information.
This is in line with the development goal of Inavi Systems to provide data to autonomous vehicles in a friendly manner.
While HD Maps are originally composed of complex shapes and various properties, requiring complex algorithms for vehicles to calculate, they are provided in the form of simple points as compressed autonomous driving information limited to a path called Trajectory.
The second is to assign a Lane Score to determine the priority of the drivable lane.
When making a left turn, the autonomous vehicle is guided to drive in the first lane as much as possible, and when making a right turn, the Lane Score is given to the rightmost lane.
The third is lane change, which generates the optimal lane change point for a turn through the actual vehicle driving trajectory.
The logic was designed to guide the lane quickly after the first turn, but it is being developed to analyze people's lane-changing patterns and improve them based on the understanding that the timing of vehicle changes differs depending on various road conditions and situations.
In response to a viewer’s question about how Trajectory can be effectively utilized to generate optimal lane change points, Director Park Jin-geun said, “We decided that it would be difficult to analyze all road conditions with logic alone.”quo; He said, “We are preparing to improve it by matching road and lane change patterns by observing the driving trajectories of actual users and testers.”
In response to a viewer's question about whether there is any development of functions to indicate road conditions (unpaved, gravel road) in addition to the navigation function, he answered, "Basically, there is a part that configures the above in the network data properties," and "Since parts such as potholes can be a risk factor for autonomous driving, we configure risk factors as a layer in the Safety Map and play a role in notifying autonomous vehicles."
