Junior serves as the primary sensing platform for the Minion Fleet and carries the team's most extensive sensor suite. Its perception and localization systems generate the information needed for environmental awareness, object identification, navigation, and coordination with the team's other vehicles.
Junior uses three Livox Horizon LiDAR units to generate dense three-dimensional point clouds of the surrounding environment. The LiDAR data is processed using the team's GB-CACHE algorithm for object extraction and classification.
GB-CACHE uses the LiDAR point clouds to identify objects in the environment and provide detections for Junior's autonomy system. These detections contribute to the vehicle's world map and support path planning and task execution.
Junior also uses a 65° field-of-view high-dynamic-range (HDR) camera for object color classification. The camera uses a polarized filter to improve color vision in the bright sunlight and water glare expected during competition in Singapore.
The LiDAR and camera systems provide complementary sensing capabilities. LiDAR supplies detailed three-dimensional information and identifies regions of interest, while the camera provides color information for object classification. The LiDAR point clouds can also provide sufficient information for object identification when camera performance is limited by sunlight or glare.
Junior uses a JAVAD DELTA-DI GNSS/IMU to provide detailed localization. This information allows sensor observations to be tied together into a common global world map.
The sensor suite follows a perception pipeline in which LiDAR detections, camera classifications, and localization data are combined to build Junior's understanding of the surrounding environment. Obstacles and identified objects continuously populate the world map, while path planning and tasking systems use this information to determine and execute vehicle actions.
Junior's software architecture uses ROS 2 and its Data Distribution Service (DDS) to allow the different perception, localization, planning, and control processes to exchange the information required for vehicle operation.
Junior serves as the surface platform and central authority within the Minion Fleet. Its primary contribution to the multi-vehicle system is environmental perception, world-map generation, and tasking while supporting coordination with the UAV and UUV.
The vehicle also includes a modular water-delivery capability as part of its available surface-vehicle functions. Junior's sensor suite was modified for RobotX 2026 to improve perception performance in the environmental conditions expected in Singapore. The combination of an HDR camera and polarized filter addresses the strong sunlight and water glare that can reduce the usefulness of conventional camera images.
The triple-Livox-Horizon configuration provides dense point clouds that support environmental perception and object identification. Even when camera data is degraded by environmental conditions, the LiDAR point clouds can provide enough information to identify objects based on their profiles.
The sensor systems are integrated with Junior's mechanical, electrical, and software architectures so that sensor data can be processed together and used throughout the vehicle's autonomy pipeline. This configuration supports Junior's role as the primary perception and tasking platform within the Minion Fleet.
