Automated robot systems that are able to navigation in remote and challenging environments like alpine regions can be a significant help for end users like mountain rescuer or constructors or maintainer of infrastructure (e.g paths or installation for protection). Such robots can perform automated transport of materials, tools, and persons as well as the automated execution of construction or maintenance actions (although the development of the actual manipulation actions is beyond the scope of this project). Endurance and payload issues renders the deployment of ground robots more realistically
Two main issues arise when deploying such robots in remote and difficult terrain. First in contrast to humans robots usually need a rather detailed maps of their environment to perform navigation; deriving and execution an efficient and safe path to a given goal. In engineered environments like urban areas or highways such detailed maps are available. In remote or unstructured areas these maps needed to be generated beforehand either by the robot itself or by other means like airborne systems.
The fact that detailed maps for remote areas do not exist and the expensive pre-mapping step are obstacle for fast and efficient deployment. Second humans are capable to navigate in a challenging environment even with a less detailed map and a rough given route because of their superior perception and motion skills. Robots in contrat still need very detailed map and a high accuracy in their localization to perform challenging navigation task.
In order to allow a ground robot to navigate efficiently and safely in remote areas to support end use activities RoboNav will aim at three main goals.
First in close cooperation with the end users use cases will be defined that are relevant for the users but are also realistic and helpful for the participating users. The definition of promising use cases and realistic requirements will maintain realistic expectations and acceptance by the users.
The second goal of RoboNav is the development of a pipeline to convert earth observation data into routing data that can be used for the navigation. The important goal is here to avoid extensive preparation campaigns such as detailed mapping of the environment with the robot system itself or other systems like UAVs. An important innovation is here that the obtained routing data will be generated depending on the robot's locomotion capabilities in order to allow broad and easy application of the approach.
The third goal of RoboNav is to develop an integrated navigation concept that is suitable for automated navigation of a robot in the envisioned challenging environments. In order to achieve this goal the views and competences of two research disciplines need to be combined: geodesy and robotics. A suitable navigation concept including localization, routing and guidance will be replicated for the purpose of an automatically moving robot.
The proposed navigation system will be implemented and integrated into a robot platform demonstrator. The integrated system will be evaluated in realistic field trials defined in cooperation with the end users. Thus, the project will be the basis for future deployments in the field but will also form a base for an economic exploitation by young participating companies from both disciplines.
Graz University of Technology
- Federal Ministry of Defense
- PS mapping services OG
- Disaster Competence Network Austria - competence network for disaster prevention
- ARTI - Autonomous Robot Technology GmbH