U.S. Military’s Top-Secret Tunnel-Crawling Robot Revealed – Part 2

Autonomously Avoids Obstacles and Creates Tunnel Maps

The Autonomous Tunnel Exploitation (ATE) robot performs initial exploration and reconnaissance missions in natural caves without GPS signals or artificial hazard facilities without prior internal information. During peacetime, it detects humans and objects in collapsed buildings during disasters.

Notably, the ATE robot demonstrated in the U.S. mainland was jointly developed by South Korea and the U.S. The Agency for Defense Development (ADD) partnered with the U.S. to secure autonomous tunnel exploration technology to explore regions without geographical information, detect hazards using robots, and create 3D maps of the exploration area. After the successful demonstration, the ATE robot, which has been developed since 2019, was assessed to have essentially reached the completion stage of development.

ADD also officially announced in June 2022 that it secured this ATE technology. ADD explained that it secured autonomous tunnel exploration technology through international joint research with the U.S. Ground Vehicle System Center (GVSC) from 2019 to 2022.

Previously, ADD introduced this ATE robot prototype at the Anheung Comprehensive Testing Site in Taean County, South Chungcheong Province, in August 2020, at the Unmanned System World Congress on September 26, 2021, and at the Seoul International Aerospace & Defense Exhibition (Seoul ADEX) in October 2021.

The ATE robot is small, measuring 1m in length, 0.68m in width, and 0.75m in height. It weighs 140kg and can maneuver at a maximum 10km/h speed.

It autonomously drives and explores the area, merging information from multiple robots to create a 3D map. It can detect and recognize pre-informed objects of interest and deploy communication repeaters strategically to extend the communication-possible area.

A drive system that can be switched between tracks and wheels depending on the situation and environment has been applied for quick maneuvering and smooth obstacle avoidance in various environments, such as natural caves and disaster areas. It is equipped with various sensors, such as LiDAR (a device that precisely draws the surrounding appearance using lasers) and CMOS image sensors.

It is also equipped with various advanced technologies, such as Simultaneous Localization and Mapping (SLAM), cartography, deep reinforcement learning-based autonomous exploration, deep neural network-based recognition, and standard hardware and software architecture.

ADD also confirmed the technical performance of the ATE robot in the tests performed through the ATE robot, successfully recognizing various forms of hazardous objects, detecting contaminated areas, and creating 3D maps with a resolution of several centimeters. This technology, which enables autonomous driving and exploration of robots without prior information in caves and underground facilities where GPS signals cannot be received, is evaluated as a revolutionary advancement of existing technology.

To Read Part 1…
To Read Part 3…