Path planning for an unmanned vehicle
AssigneeElta Systems Ltd.
InventorsOfir Cohen, Dina Appelman
Life of this patent
Estimating the term…
Abstract
Thus, according to some examples of the presently disclosed subject matter, in order to reduce path planning time, information that has previously been calculated during planning of a path is reused during planning of other subsequent paths. Using the stored indication reduces time required for planning the path and thus enables to evaluate a greater number of optional paths within a given period of time. This can assist in increasing speed and smoothness of vehicle maneuverability.
Claims32
A method of path planning for an unmanned vehicle (UV) traversing an area, the method comprising:
- operating a scanning device for executing scanning operations, comprising scanning a region within the area, to thereby generate respective scanning output data;
- obtaining a map based on the scanning output data, the map representing at least a part of the region, the map is divided into cells, each cell is classified to a class selected from at least two classes, comprising traversable and non-traversable;
- based on the map, planning one or more paths for the UV, the planning comprising:
- upon determining that the UV is to assume a vehicle within a path of the one or more paths:
- querying a data structure;
- the data structure is configured to maintain one or more entries, each of the entries represents a respective vehicle-state characterized by at least a respective position and heading of the UV, and is indicative of whether the vehicle-state is feasible for the UV;
- if the data structure comprises an indication corresponding to the vehicle-state, using data stored in association with the vehicle-state for planning the path, otherwise computing and storing the data associated with the vehicle-state;
- and
- selecting a path from the one or more paths.
The method of any one of the preceding claims, further comprising steering the UV in accordance with the path that has been selected.
The method of any one of the preceding claims, wherein feasibility of the vehicle-state requires that position of the UV at the vehicle-state is not mapped to a non-traversable cell.
The method of any one of the preceding claims, wherein feasibility of the vehicle-state requires that yaw and pitch angles of the UV comply with one or more predefined criteria.
The method of any one of the preceding claims, wherein the at least one entry further comprises a weight indicating proximity of the UV to an obstacle when the UV assumes the vehicle-state.
The method of any one of the preceding claims, wherein the at least one entry is maintained for further maps generated upon receiving further scanning output data.
The method of any one of the preceding claims further comprising operating the processor for generating the map.
The method of any one of the preceding claims, wherein the unmanned vehicle is an unmanned ground vehicle.
The method of any one of the preceding claims, further comprising receiving data indicative of a destination of the UV, and wherein the one or more paths are determined from the current position of the UV to the destination.
The method of any one of the preceding claims further comprising:
- as the UV advances within the area and new scanning output data of new regions within the area is generated, obtaining an updated map generated and based on the new scanning output data and repeating the planning based on the updated map.
A system mountable on an unmanned vehicle (UV), comprising:
- a processing unit configured and operable to:
- receive scanning output data, the scanning output data is a product of scanning operations executed by a scanning device on-board the UV and include scanning a region within the area;
- obtain a map based on the scanning output data, the map representing at least a segment of the region, the map is divided into cells, each cell is classified to a class selected from at least two classes, comprising traversable and non-traversable;
- based on the map, planning at least one path for the UV, the planning comprising:
- upon determining that within one of the at least one path, the UV is to assume a vehicle-state:
- query a data structure configured to keep one or more entries, each of the entries indicating whether a respective vehicle-state is feasible for the UV, the vehicle-state being associated with at least a respective position and heading of the UV;
- if the data structure comprises an indication corresponding to the vehicle-state, using data stored in association with the vehicle-state for planning the path, otherwise computing and storing the data associated with the vehicle-state;
- and
- select a path for the UV.
An unmanned vehicle (UV), comprising:
- a scanning device and a processing unit comprising one or more processing circuitries;
- the scanning device is configured and operable to scan a region surrounding the UV within an area traversed by the UV to thereby provide scanning output data providing information about distances between objects in the region and the UV in a multiplicity of directions;
- the processing unit is configured and operable to:
- generate a map based on the scanning output data, the map representing at least a part of the region, the map is divided into cells, each cell is classified to a class selected from two or more classes, comprising traversable and non-traversable;
- receiving data indicative of a destination of the UV;
- and, based on the map, planning one or more paths to a destination for of the UV,
- the planning comprising:
- upon determining that within a path of the one or more paths the UV is to assume a vehicle-state:
- querying a data structure, stored in a computer storage device operatively connected to the processing unit, configured to maintain information indicating whether a respective vehicle-state is feasible for the UV, the vehicle-state being associated with at least a respective position and heading of the UV;
- if the data structure comprises an indication corresponding to the vehicle-state, using data stored in association with the vehicle-state for planning the path, otherwise computing and storing the data associated with the vehicle-state;
- and selecting a path from the one or more paths.
A non-transitory computer readable storage medium retaining a program of instructions, the program of instructions when read by a computer, cause the computer to perform a method of path planning for an unmanned vehicle (UV), comprising:
- obtaining a map based on scanning output data;
- the scanning output data is a product of scanning operations executed by a scanning device on-board the UV, during which a region within an area traversed by the UV is scanned;
- the map is divided into cells, each cell is classified to a class selected from two or more classes, comprising traversable and non-traversable;
- based on the map, planning two or more paths to a destination of the UV, the planning comprising:
- upon determining that within a path of the two or more paths, the UV is to assume a vehicle-state:
- query a data structure configured to maintain information indicating whether a respective vehicle-state is feasible for the UV, the vehicle-state being associated with at least a respective position and heading of the UV;
- if the data structure comprises an indication corresponding to the vehicle-state, using data stored in association with the vehicle-state for planning the path, otherwise computing and storing the data associated with the vehicle-state;
- and selecting a path from the two or more paths.
Dependencies are inferred from the claim text.