Team
Our team at ESE day!
Background Information
We discussed with the National Geospatial-Intelligence Agency (NGA) to identify challenges relating to geospatial information. From this discussion, our focus became effectively allocating tasks across various robotic platforms while correlating satellite data with ground-level information. This project addresses the limitations of 2D overhead satellite data such as MAXAR and determines the challenges associated with using this data for robotic navigation. Further, the NGA aims to leverage this data to optimize the allocation of autonomous robot platforms to specific areas of interest. For instance, they seek to identify areas with obstacles that may hinder the operation of ground-level robots, prompting the deployment of alternative platforms such as robotic dogs.
Our project takes an initial step towards addressing this challenge by harnessing open-source satellite street data from OpenStreetMap to enable semi-autonomous navigation for a drone.
Problem Statement
How can robotic navigation enhance geospatial information collection, specifically addressing 2D overhead satellite imagery limitations?
Current geospatial information collection methods, particularly relying on 2D overhead satellite imagery, face limitations in accurately capturing ground-level information and navigating complex environments. This limits understanding and analysis of areas such as the East End at Washington University in St. Louis, where traditional satellite imagery may not suffice to provide detailed insights into conditions on the ground.
This project, conducted in collaboration with the National Geospatial-Intelligence Agency (NGA), focuses on integrating satellite data with ground-level information obtained via drones. By leveraging OpenStreetMap (OSM) data and robotic navigation, the project aims to overcome imperfections in semi-autonomous route traversal. The specific scope of the project involves using drones to capture images and detect obstacles along sidewalks within the East End at Washington University in St. Louis. These observations are then uploaded to the NGA’s Mobile Awareness GEOINT Environment (MAGE) software for analysis. The ultimate goal is to produce a detailed report containing captured images and route information, demonstrating the potential of robotics in enhancing geospatial applications and navigating complex environments inaccessible to traditional overhead satellites.
Image of OSM
Project Objectives
We aim to use robotic navigation to overcome imperfections in semi-autonomous route traversal. We extract path information from OSM to define routes for the drone to follow. The sidewalks within the East End at Washington University in St. Louis will be used as the area scope for this project. The drone detects obstacles and captures images at each specified path crossing (node), providing comprehensive visual data. More specific project objectives are outlined below.
- Integrate satellite data with ground-level information obtained via drone to enhance geospatial information collection.
- Address imperfections in semi-autonomous route traversal by utilizing robotic navigation techniques.
- Overcome limitations of 2D overhead satellite imagery by leveraging OpenStreetMap (OSM) data to define routes for drones.
- Capture comprehensive visual data along specified paths within the East End of WashU’s campus, focusing on sidewalk areas, using drones equipped with obstacle detection capabilities.
- Upload observations to the National Geospatial-Intelligence Agency’s (NGA) Mobile Awareness GEOINT Environment (MAGE) software for analysis.
- Produce a detailed report containing captured images and route information, showcasing the knowledge gained through high-level data integration and low-level observations.
- Demonstrate the potential of robotics in enhancing geospatial applications and navigating complex environments that are unreachable by existing overhead satellites.