# AI-Powered Vibration Analysis Enhances Autonomous Robotic Subsurface Exploration New research enables robots to autonomously determine the burial depth of rocks using vibrations. - **Source:** [Universe Today](https://www.universetoday.com) - **Published:** 2026-11-09 - **Topic:** [science](https://brifing.si/c/science) - **Original article:** - **On Brifing:** https://brifing.si/s/f1ae527a-dd89-445f-9a90-dfefefc331a4 ## Autonomous Subsurface Mapping Planning robotic missions requires precise knowledge of surface and subsurface terrain. Before landing, mission controllers must understand if obstacles are movable or fixed. Virginia Tech researchers developed a new tool allowing robots to autonomously determine the burial depth of rocks, significantly improving the efficiency of future planetary exploration and excavation tasks. ## Vibration-Based Detection Mechanism The research team, led by Yiyan Ruan and Erik Komendera, utilizes vibration analysis to map underground structures. By striking exposed rock portions and measuring resulting vibrations, the system gathers critical data. This method provides a reliable way for machines to assess the depth of partially embedded rocks without human intervention or oversight. ## Neural Network Integration The team employed a Multilayer Perceptron, a specific type of neural network, to interpret complex vibrational data. By training the model on eleven measurable features, the system successfully identifies non-linear patterns. This compact neural network, featuring only two hidden layers, effectively correlates impulse excitation with rock geometry and actual burial depth. ## Experimental Validation Process To validate their approach, researchers buried eighteen concrete blocks of varying sizes and shapes within sand at different depths. They utilized an instrumented hammer to strike the blocks and a Laser Doppler Vibrometer to record the data. These controlled experiments provided the necessary foundation for training the AI model to recognize depth. ## Energy Efficiency for Rovers Robotic rovers on Mars operate under strict energy budgets, making obstacle avoidance or removal critical. Knowing whether a rock is shallow or deeply embedded prevents rovers from wasting power on impossible tasks. This autonomous capability allows machines to make real-time decisions, ensuring safer navigation and more effective science operations on planetary surfaces. ## Overcoming Complex Variables Estimating burial depth is inherently difficult due to variations in rock geometry and soil composition. The researchers successfully addressed these challenges by carefully selecting features from vibration measurements. Their model accounts for how soil holds a rock, producing reliable data despite the complex environmental factors that typically complicate subsurface structural analysis. ## Terrestrial and Industrial Applications Beyond space exploration, this technology offers significant potential for applications on Earth. The team plans to test the approach for analyzing construction sites and identifying hazardous materials buried underground. Furthermore, the system could assist in studying unstable rock formations and underground caverns, aiding in resource planning and disaster management efforts globally. ## Future Exploration Implications This development marks a shift toward more autonomous robotic capabilities in challenging environments. By enabling machines to interpret their surroundings, future missions can better locate water ice and navigate difficult terrain. The ongoing research aims to refine these datasets, ultimately advancing our ability to explore both planetary surfaces and Earth's subsurface. ## Sources covering this story - https://www.universetoday.com/articles/virginia-tech-researchers-use-vibrations-and-ai-to-estimate-depth-of-buried-objects-for-future-lunar — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._