Embedded Files
A-TARE planner is the aerial version of our exploration planner. The planner involves a hierarchical framework for highly efficient exploration - one level in the framework maintains data densely and computes a detailed path within a local planning horizon, another level maintains data sparsely and computes a coarse path at the global scale. The paths at both levels are joined together to form the exploration path. The framework draws the insight that detailed processing is most effective close to the vehicle, while coarse processing provides sufficient utility far away from the vehicle. The framework trades-off details at the global scale for computational speed. The local path is piecewise smoothed in path segments, suitable for aerial vehicles to execute. The aerial vehicle follows a path segment to the end, then stops and turns before following the next path segment.
A-TARE Planner will be commercially available in the near future.
A-TARE hierarchical exploration framework. Inside the local planning horizon, data is densely maintained and a local detailed path (dark-blue) is computed. At the global scale, data is sparsely maintained in the distant subspaces and a global coarse path (light-blue) is computed. The local path and global path are connected on the boundary of the local planning horizon to form the exploration path.
The local path is smooth in path segments. Viewpoints on a path segment (solid orange dots) are points where the vehicle passes without stopping, and viewpoints between two path segments (hollow orange dots) are points where the vehicle stops and turns.
Results
Results
The methods are evaluated in the campus environment in Autonomous Exploration Development Environment. The vehicle navigates at 5m/s. Each method is run a number of 10 times. The trajectories are the best of the 10 runs and the exploration metrics (explored volume, traveling distance, and algorithm runtime on vertical axis) are the means of the 10 runs with the time on horizontal axis spanning the longest run. A run is ended if the method reports completion, the vehicle almost stops (< 10m of movement within 5 minutes), or the time limit is reached. Here, the time limit is set to four times of A-TARE Planner. The algorithm runtime is evaluated based on a 4.1GHz i7 CPU in system time. All algorithms use a single CPU thread for exploration planning.
Baseline methods
NBVP: Bircher et al. Receding Horizon “Next-Best-View” Planner for 3D Exploration. ICRA 2016.
MBP: Dharmadhikari et al. Motion primitives-based path planning for fast and agile exploration using aerial robots. ICRA 2020.
GBP: Dang et al. Graph‐based subterranean exploration path planning using aerial and legged robots. Journal of Field Robotics. 37(8):1363-1388, 2020.
Campus Environment
Campus Environment
References
References
C. Cao, H. Zhu, Z. Ren, H. Choset, and J. Zhang. Representation Granularity Enables Time-Efficient Autonomous Exploration in Large, Complex Worlds. Science Robotics. vol. 8, no. 80, 2023. [PDF] [Summary Video]
C. Chao, H. Zhu, H. Choset, and J. Zhang. TARE: A Hierarchical Framework for Efficiently Exploring Complex 3D Environments. Robotics: Science and Systems Conference (RSS). Virtual, July 2021. Best Paper Award and Best System Paper Award. [PDF] [Spotlight Talk]
C. Chao, H. Zhu, H. Choset, and J. Zhang. Exploring Large and Complex Environments Fast and Efficiently. IEEE Intl. Conf. on Robotics and Automation (ICRA). Xian, China, May 2021. [PDF] [Talk]
People
People
Chao Cao
CMU Robotics Institute
Hongbiao Zhu
CMU Robotics Institute
Howie Choset
CMU Robotics Institute
Ji Zhang
CMU NREC & Robotics Institute
Notes
Notes
A-TARE planner is developed using Aerial Navigation Development Environment, while both are being commercialized by the National Robotics Engineering Center (NREC) of CMU.
TARE is named after the efforts to develop Technologies for Autonomous Robot Exploration.
Page updated
Google Sites
Report abuse