| Verification of the Feasibility of Mixed Reality for Test and Evaluation of an Autonomous Platform in Off-road Environment |
| Gyuhwan Hwang, Saedong Yeo, Hyunsung Tae |
| 6th R&D Institute, Agency for Defense Development, Daejeon, 34186, Republic of Korea |
|
|
|
Received: November 13, 2024; Revised: April 23, 2025 Accepted: July 7, 2025. Published online: July 25, 2025. |
|
|
|
| ABSTRACT |
|
Applying mixed reality in test and evaluation of an autonomous system is promising, because it can narrow the gap between the live test and virtual test. In this study, we verified the feasibility of mixed reality in test and evaluation of an autonomous platform in off-road environment. We developed a scheme for mixed reality before implementing it for a real platform and environment. The verification of the scheme was performed on a virtual prototype test bed which consists of two virtual spaces in a single Gazebo simulation environment. In one space, the real vehicle and real environment are modeled, and in the other space, sensor signals for a virtual object are generated. The signals from two spaces are mixed as an input to a perception module of the autonomous platform. The effectiveness of the test bed was validated both qualitatively and quantitatively by checking whether the sensor signals for a virtual object can be generated in the correct position and pose at the correct time, as intended. We also demonstrated that the scheme for mixed reality developed on the virtual prototype test bed can be directly applied to real vehicles and actual driving environments. |
| Key Words:
Autonomous driving · Mixed reality · Real time · Simulation · Test bed · ROS · Gazebo |
|