Mapping on Demand

• Applicant: Professor Dr. Cyrill Stachniss
• Subject Area: Geodesy, Photogrammetry, Remote Sensing, Geoinformatics, Cartography
• Term: from 2011 to 2019
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 166047863

As in the first phase of this project, the goal is to investigate, develop, and integrate new methods that enable fast, effective, and autonomous acquisition and interpretation of 3D scenes based on a user query. The target platform is a light weight autonomous multicopter. The copter uses visual information from cameras and laser range data to perceive its surroundings, to model the scene, to navigate, and explore the environment. The research unit develops and extends the required models, methods, and techniques to achieve these goals and will evaluate the prototypes in residential suburb-like environments. The methods to be develop focuses on the autonomous acquisition of probabilistic models to capture spatial-temporal patterns including semantics. The ability to cope with noisy sensor data and to explicitly represent the uncertainty is a central design element. The targeted approaches can be summarized with the term Mapping on Demand. This includes all processes, techniques, and tools to acquire the sensor data, to process and to interpret it with the goal of building a semantically annotated 3D model for the user. The user obtains the model on time and it supports his decision making process. Within the second phase, the research unit focuses on techniques for onboard execution and that are of incremental nature. The user of the proposed mapping and exploration system will not control the copter in the way it is done today, as a pilot with a remote control. Instead, the copter autonomously executes a specified mission without requiring a pilot on site. In this sense, the project goes beyond traditional approaches as it requires localization, mapping, exploration and navigation at the same time. Therefore, the research unit targets on integrated approaches that can address all problems in a joint manner, taking into account the inherent noise and uncertainty. A further challenge arises from the light weight micro aerial vehicle with an overall weight of less than 5kg. In the first phase of Mapping on Demand, we developed - in line with the proposal - several components in a real-time design and evaluated the prototypes using single-flight missions without temporal constraints or substantial changes in the environment. In the second phase, we will extend the existing components and develop new methods for incremental operation on the copter. Among other topics, we will * extend and robustify the obstacle detection by considering more sensor modalities in an integrated fashion together with the semantic interpretation,* increase the precision and reliability of the trajectory estimation processes considering also sensor failures,* speed up the 3D surface reconstruction and scene interpretation so that it can be executed during the flight for supporting the navigation system, and* develop novel exploration approaches that consider background knowledge allowing more flexible missions.

• DFG Programme: Research Units
• Subproject of: FOR 1505: Mapping on Demand