Patent Application
20210114622
This application claims priority to Finnish Application No. 20195897, filed Oct. 17, 2019, the entire contents of which are incorporated herein by reference.
The present disclosure relates to methods, apparatuses and computer program products for movement control of a device.
A vehicle or other movable device may operate autonomously or semi-autonomously based on control instructions defining movement between two locations. Control instructions may relate to various aspects of a path of movement the device shall follow. For example, instructions regarding turns, altitudes, speed, acceleration, breaking, obstacles and locations to avoid and so on may be provided. The control instructions can be provided by a control apparatus. The control apparatus can be remote. Alternatively, or in addition, at least a part of the control instructions may be provided by on-board processor apparatus. Control instructions may be provided from a remote apparatus for an unmanned device, for example, via an appropriate data communication system, for example over a wireless link.
An autonomous device can be defined as a device that is not under direct control of a human operator. An example of autonomous devices are unmanned vehicles, such as aerial vehicles (e.g. unmanned aerial vehicles (UAVs); often referred to as drones), land vehicles and watercraft and other vessels. Non-limiting examples of land vehicles comprise moving vehicles such as automotive (e.g., driverless cars, vans, heavy good vehicles, motorcycles etc.), industrial automatic guided vehicles, farming, forestry, gardening, cleaning, clearing, and surveillance and/or local control equipment and so forth. Further non-limiting examples of unmanned moving devices comprise machines such as robots, manipulators and other machines that can move in an area or space without need of direct control by a human operator.
Movement of a device can be controlled remotely by sending via a communication link control instructions from a separate control apparatus defining a path of travel to be followed. At least part of the processing may take place at the moving device.
An area where the device may move and needs to be controlled may be large. Processing the necessary information may require considerable data processing and/or memory capacity. Controlling paths of movement may cause latency and/or overhead issues.
According to an aspect, there is provided an apparatus comprising data processor means configured to: determine an initial path past an obstacle between a first point and a second point, wherein the initial path comprises at least two turn points between the first point and the second point, determine whether a line between two of the points and bypassing at least one of the turn points intersects the obstacle, and determine a straightened path of travel in response to determination that the line between two of the points and bypassing at least one of the turn points does not intersect the obstacle by removing the at least one bypassed turn point from the initial path.
The data processor means can be configured to determine the initial path based on a grid. The data processor means may be configured to use grid cells and/or real-world coordinates in determining whether a line bypassing at least one of the turn points intersects the obstacle. Conversion between the grid and real world coordinate systems may be provided.
The data processor means may be configured to determine at least one local grid covering at least one obstacle, and to determine a local initial path past the at least one obstacle using the local grid. The data processor means may be configured to determine a local grid covering at least two obstacles. The data processor means may be configured to determine a plurality of local grids for a search path between start and target locations, determine a straightened path for the plurality of local grids, and use the straightened paths in generating a path for movement from start to target.
The data processor means may be configured to repeat the determining whether a line between two points and bypassing at least one turn point intersects with the obstacle until there are no more turn points to be tested.
The data processor means may be configured to enlarge a grid covering at least one obstacle by means of padding.
The data processor means may be configured to determine zoom patterns by determining at least one zoom pattern that has at least one obstacle to movement, dividing the determined at least one zoom pattern into smaller zoom patterns, determining at least one of the smaller zoom patterns with at least one obstacle, repeating the dividing step until predefined smallest zoom pattern size is reached. A path generated using the at least one determined zoom pattern may then be straightened. The data processor means may be further configured to generate a path of travel between a start point and an target point based on a combination of information of the straightened path and at least one search path outside the at least one determined zoom pattern. The data processor means may be configured to apply a path finding algorithm only to zoom patterns that have been determined to have at least one obstacle and assume that there are no obstacles in areas outside the determined zoom patterns.
The apparatus can be comprised in one of an unmanned aerial vehicle, unmanned land vehicle or unmanned vessel. Means for receiving control instructions from a remote station may be provided. The apparatus may comprise on-board data processing means configured to provide at least a part of the determinations.
The apparatus may comprise a ground control station.
In accordance with an aspect there is provided a method of determining travel between a first point and a second point, the method comprising: determining an initial path past an obstacle between the first point and the second point, the initial path comprising at least two turn points between the first point and the second point; and in response to determining that a line between two of the points bypassing at least one of the turn points does not intersect the obstacle, removing the at least one turn point from the initial path to determine a straightened path of travel.
The determining of the initial path may comprises using a grid on the area of travel. The determining whether the line between the two of the points bypassing at least one of the turn points intersects the obstacle may comprise using grid cells and/or real-world coordinates of the area. A conversion between grid cells and real world coordinates may be provided.
At least one local grid on may be determined on an area around at least one obstacle. A separate initial path of travel past the at least one obstacle can be determined using the local grid. At least two obstacles can be covered by a single local grid.
The method may comprise repeatedly determining if a line between two points bypassing at least one intermediate turn point intersects with the obstacle until there are no more points to be tested.
The method may comprise padding a grid covering at least one obstacle.
In accordance with an aspect a method comprising configuring patterns covering sectors of a controlled area, the configuring comprising determining at least one pattern that has at least one obstacle to movement in the area, dividing the determined at least one pattern into smaller patterns, determining at least one of the smaller patterns with at least one obstacle to movement in the area, repeating the dividing step until predefined smallest pattern size is reached can be applied in combination with the grid based pathfinding and/or path straightening methods.
Determined travel between a first point and a second point can be defined by a determined at least one pattern, the method comprising generating a path of movement for a device based on a combination of information of the determined travel between the first point and the second point and a search path outside the at least one pattern.
The path finding methods may only be applied to patterns that have been determined to have at least one obstacle to movement. Any path of movement in an area outside the determined patterns can be generated based on assumption that there are no obstacles to movement within the area outside the determined patterns.
A computer software product embodying at least a part of the herein described functions may also be provided.
Some aspects will now be described in further detail, by way of example only, with reference to the following examples and accompanying drawings, in which:
In general, the following detailed description is given with reference to movable devices such as vehicles that operate based on control instructions received from a remote control apparatus. It is, however, noted that although the detailed examples are given in the context of unmanned vehicles, or autonomous vehicles receiving control instructions from separate control apparatuses, the determinations and computations may also be provided partially or entirely by control apparatus provided on the moving device itself.
The examples relate to control instructions for paths of travel, in particular instructions to avoid prohibited zones such as no-fly zones (NFZ) and other no go areas. Such areas can be referred to as items relevant to movement or simply as obstacles. Different methods may be used to generate path between the starting and target points wherein at least one prohibited zone may occur in the path. For example, a path planning system may be based on use of a grid structure to cover whole path area, a grid created using zooming tiles, or a grid created around the prohibited zone(s), in order to avoid travelling through a prohibited zone. In the following various alternative solutions are provided. There are different benefits to choose one or more of these when generating the path. The whole path may have segments or sectors which are created by different ways, for example.
The control apparatus 10 can comprise at least one processor 11, 12 and at least one memory 15. The at least one memory 15 comprises computer code that, when executed on the at least one processor, causes the apparatus to perform one or more of the herein described functions. The control apparatus 10 can be configured to communicate via appropriate data communication system using appropriate one or more communication protocols. The communications may be via local networks, wide area networks or even direct communications between the control station and the unmanned device. For example, communication may be based on 4th or 5th generation (4G, 5G) communication systems and protocols, or any later developments of communication systems. The communications may be carried at least in part on wireless links 24. The protocols may be based an appropriate connectionless protocol. Thus the remote control station 10 is capable of sending messages to an on-board data processing apparatus of the unmanned vehicle 20. The control station may also be configured to receive messages from the unmanned vehicle. The control apparatus can comprise data communications circuitry, denoted by reference 14, for receiving and transmitting data. It is understood that although the communications circuitry and various possible components thereof are shown as one block, the circuitry can comprise a number of circuitries. Such circuitries may share at least some components between them.
Instruction data 16 is shown to be available at the at least one memory 15. The instruction data can comprise control instructions regarding path of travel, for example information about location coordinates to be followed to control, e.g., at least one of longitude and latitude, altitude, speed, acceleration, breaking, distance, and so on. Control instruction items for a path of travel are often called waypoints. Examples for determining instructions regarding the path of travel to the unmanned device 20 will be described below.
The unmanned aerial vehicle (UAV) 20 is configured to receive control information from the control station 10. The unmanned aerial vehicle of the example of
The memory 28 can provide a data buffering function for control instruction data 29. The at least one data processor can read the control instructions from the buffer and cause performance of operations relating to the task to be performed accordingly. The control apparatus may be configured to provide on-board data processing apparatus. The on-board data processing apparatus of the unmanned vehicle can generate autopilot-specific instructions from the received commands.
The control station 10 may receive telemetry information from unmanned vehicles under its control. For example, an unmanned vehicle may be configured to transmit mission progress information (e.g. information about the current waypoint and remaining distance, about near-by objects, fleet, other moving vehicles, other moving vehicles in a swarm of vehicles, sensor data from the device, and/or other devices and so on) to the ground control station. Information about the operating condition and/or state of the device may also be provided. For example, information about the remaining energy levels may be communicated back to the control station. The control station can then take the received information into account in the control actions, including in decision making regarding what to include in messages to the unmanned vehicle.
The following describes certain more detailed examples of determining movement paths for unmanned devices. Optimization of usage of computer resources such as processing power and dynamic memory for determining paths of movement are also described. The examples illustrate principles that can be applied to any unmanned vehicle or the like. For the purposes of illustration however the exemplifying device is specified to comprise an unmanned aerial device (UAV) and control apparatus adapted for sending of control data is referred to as a ground control station (GCS). A ground control station (GSC) computer can be configured to control one or more UAVs flying over large areas. GSC computer may comprise various features such as input and output equipment, display, keyboard, mouse, touchscreen, and so forth. The area controlled by a control station may be substantially large, for example 100×100 km or even larger.
Movement of unmanned moving vehicles such as aircraft, e.g. drones, land vehicles, watercraft, robots, and the like can be controlled based on path finding algorithms. For example, a control station such as a GCS can use specific path finding algorithms to create a safe and allowed path for a drone mission over an area. Path finding algorithms are configured to circumvent obstacles or no-go areas such as “Non Fly Zones” (NFZ) on the path. It is possible that operation area of, e.g., a drone in a GCS control area may have several Non Fly Zones. These areas can be closed for various reason for flights over them, or flights on some altitudes over them. NFZ size can be anything from a few meters to some hundreds of kilometres and can be of any shape. A controlled area may have any amount and distribution of NFZs. A GCS can use a path finding algorithm to generate optimal route for a drone from location A to location B so that crossing of any NFZs is avoided. NFZs are predefined in the maps data or applications used by GCS and they can be found in real-world coordinates as polygons.
In an example UAVs such as drones can navigate around no-fly zones using a two dimensional (2D) grid-based path finding algorithms. A non-limiting example of a possible path finding algorithm is a “Jump Point” path finding algorithm. “Jump Point” algorithms can be used for multiple random distributed NFZs of any shapes. “Jump Point” algorithms typically use uniform grid of open and closed nodes. The grid can be 2-dimensional {latitude, longitude} or 3-dimensional {latitude, longitude, altitude}. When a drone's path requires accuracy of a few metres then a grid for a “Jump Point” algorithm needs to have a step length of a few meters. This can consume substantial amount of memory capacity. For example, grid with a 2 m step for an area 2000×2000 m takes 1 Mbyte of computer dynamic memory in 2-dimensional (2D) case and 100 Mbytes for 3-dimensional (3D) case, assuming typical altitude range from 0 to 300 m. For an area of 100 km×100 km 250 MB for 2D and respectively 25 GB for 3D would be needed. That may exceed available memory resources of a GCS computer and make the algorithm slow, especially on a large grid.
A two-dimensional path finding grid can be created on top of a real-world plain area map and presented on a display of ground control station or other computer. The initial inputs in a grid can include the drone/UAV coordinates, target coordinates, and no-fly zone(s) coordinates. The grid is configured to define drone location, target location and no-fly zone(s). Grid cells touching a no-fly zone can be marked as obstacles in the grid. A 2D grid based path finding algorithm (for example a Jump-Point Search) can be applied to a grid to generate an initial path of travel.
In further example embodiments cells are using middle point of each cell as potential waypoint of the path. If the middle point of cell of grid is covered by obstacle, cell is blocked. Blocked means that the cell cannot be used as part of path in grid based path finding algorithm, for example. Otherwise the grid can be used as free cell and can be used in the path. The path calculated follows the middle point of selected free cells as waypoints around the obstacles. The use of grid avoids the calculations of added dot markings on the display. The number of dot markings may be not exact and in case of many dots it might make the calculation problematic in digitalized environment and accuracy is decreasing.
Furthermore, in example embodiment there are two points of initial path which intersect with outer grid area around the obstacle. The first intersection point is in that side of grid which is directing the start of path and second intersection point in that side of grid which is towards end of path. The intersection points corresponding cells of grid, respectively, are selected as start and end points of the path search algorithm like grid based path finding algorithm. The algorithm is run.
In example embodiment when the waypoints are in consecutive order, first three waypoints may be used, for example. First three waypoints may be started from different points. Intermediate (center) waypoint is located between first and third waypoints of the three consecutive waypoints. Depending on the situation the intermediate waypoint may be used in the path or it can be removed. Intersection analysis may be continued or repeated with next selected points until there are no further removal in the selected three consecutive points, for example.
In example embodiment after the path between two points of initial path which intersect with outer grid area around the obstacle is finished, the path obtained from grid based path finding algorithm is combined with the initial path to compose the final path.
The path cell coordinates can be converted to real-world coordinates and other way around. Conversion between real world map coordinates and computational grids as such is a feature of mapping and pathfinding products. Thus real world x and y coordinates can be used in a path finding algorithm and conversion provided to and from latitude and longitude using a grid based system which has a conversion functionality.
In accordance with the herein disclosed principles the initial path is cleaned-up from excess waypoints by looping through path waypoints. Examples of the clean-up will be described in more detail below.
In case of a long distance between a start location and a target location and scattering of no-fly zones around the area, the grids can be created locally around no-fly zones that are on the direct path from start to target. Local grids can then be solved separately.
In accordance with an aspect, for multiple overlapping or close no-fly zones a shared local grid may be generated which covers the overlapping or close no-fly zones.
The grid 30 comprises cells 34. A non-limiting example of the cells size is 2 m×2 m cells for drone, for example. Cell about this size have proven to work well in practice. However, differently sized cells may be used. Small cells can be used to obtain a more accurate path. On the other hand, larger cells would give more performance. Thus optimisation of the cell size may be desired. There may also be desire to be able select the cell size based on needs of the particular use. Smaller cells provide more accuracy and larger cells will provide more performance.
Grid cells touching a no-fly zone can be marked as obstacles. In
The grid-based path finding algorithm can then be applied to the grid to find waypoints 36 (the waypoint cells are shown in light grey) around the obstacle 33. For example, a jump point search algorithm may be applied to the grid at this stage. The resulting turning cells, i.e., waypoints 36 around the obstacle 33 found by jump-point search algorithm are shown in
Obtained path cell coordinates can then be converted to real-world coordinates to provide an initial rough path 37 via the waypoints 36, this being shown in
The 2D grid based algorithm may only support movement in eight directions, and because of this the initial rough path 37 may comprise one or more excess waypoints, and thus unnecessary turns. A clean-up operation may be provided to remove any possible excess waypoints.
A clean-up algorithm may comprise looping through the waypoints 36 on the initial path 37 to check if any of these can be removed.
Instead of n being an integer, a non-integer or decimal number, for example, may also be used. The latitude and longitude can be calculated, e.g., by distance to the point (n+1)/p or (n+2)/p from n, where p can be integer. This might be case, e.g., when another NFZ is located at the other end of a route.
In
In
The looping can now progress to the step shown in
In
A problem is how to find a straightened line. If the path is calculated by grid based or tile based algorithm, for example, the determined path may not necessarily be the straightest. For straightening turn points (e.g. waypoints) calculated from the algorithms may be used together with start and target points in the path. The path may be straightened by calculating the distance from start point through first turn point to second turn point. If the line from the start point to the second point is not intersecting the obstacle, and the distance is less than the distance through the start, first and second points, the straightening may be provided and the first turn point is not needed (point n+1 in
In this example the check was made in view of n+2 waypoints. However, a check can also be made in view of a waypoint further ahead. That is, instead of having a loop where the testing is based on n+2 the looping can be based on testing n+m where m greater than 2. If there is no intersection, any intervening waypoints (n+1 to m−1) may then be ignored at once.
Line intersection checks can be done using real world coordinates instead of grid cells. If need, conversion between the real world coordinates and grid cells can be provided. Use of real world coordinates may provide some advantage, e.g., in view of better accuracy. E.g., the check at the step of
Any appropriate line intersection check algorithm can be used for the check. In computations, x and y coordinates can be replaced with latitude and longitude values. A non-limiting example of the intersection check is shown in
In accordance with a further example a long distance path is determined from start to target in environment where there are one or a number of no-fly zones scattered in the area. A problem in such circumstances may arise because of a possible need of a substantially large grid, and this may not be feasible for performance reasons. A way of addressing this is to create one or more grids or a combination of grids around the one or more no-fly zones that are on the direct path from start to target. The paths within the one or more grids can be solved as explained.
Such situation is illustrated in
The direct parts of path are usually much longer than the parts of obstacle, in drone paths, for example. One reason is that obstacles are tried to be reduced in the path planning. The paths in the air are less crowded with obstacles and more freedom is possible to determine the path, for example. The path on the ground are having more obstacles in the paths compared to air where drones are used and therefore more grid based algorithm calculation in the path on the ground is needed and also the whole path can use grid based algorithms. Drones as flying objects need power efficiency, and less computation is better enhancing flying time, for example.
Local grids 65 and 66 can then be created around relevant NFZs 62 and 63, see grids in
If the middle point of cell is covered by the obstacle, cell is blocked. Blocked means that the cell cannot be used as part of rough path waypoint in grid based algorithm. Otherwise the grid cell, if the middle point of cell is not covered by obstacle, can be used as free cell, i.e. middle point as waypoint can be used in the initial path.
As known the real-world coordinates are in a form of “latitude, longitude” or normal wgs84 coordinates. Coordinates for grids start (0, 0) from left top corner of the selected grid covering obstacle and are integers. When grid is created it may be created to some area in the world, i.e. area of the obstacle. The creation may be done so that real world coordinates corresponds grid cell (0, 0). The size of cell of grid is 5 m, for example, it describes how much the coordinate changes between the cells. For example, in a grid with cell size of 5 m, the cell (10, 5) is defined to be 50 m to the East and 25 m to the south of cell (0, 0).
Pathfinding as explained above can be applied to each grid to determine rough paths through the grids 65 and 66. This is illustrated in
In
If the line tested does intersect with the obstacle, it can be determined at 106 that the straightening is not possible. Testing can then be repeated at 108 for a next combination of points.
At the repeat stage it is possible to use the next point on the initial path as a starting point. Alternatively, if two or more intermediate points were bypassed at the previous test cycle the next test cycle may try with a lesser amount of intermediate turn points.
The determining of the initial path can comprise using a grid on the area of travel. The grid may be a local grid. The determining whether the line between two of the points passing at least one of the turn points intersects the obstacle can comprise use of the original or real-world coordinates of the area.
Multiple no-fly zones can be located such that use of local grids may produce overlapping local grids. A shared local grid may be generated that covers the two or more zones. A shared grid is illustrated in
Grid generation may result a situation illustrated in
In accordance with a further aspect further optimisation of use of local grids can be provided. In accordance with an example a controlled area can be divided in smaller geometrical patterns covering sectors of the area, the patterns comprising indication whether there are obstacles or not within the patterns. Instead of applying a path finding algorithm to the entire length of travel between locations A and B, the path finding algorithm can be applied selectively based on the indications to one or only some of the patterns in the route between locations A and B.
Examples of division of an area 110 to a multiple of patterns are shown in
The process can be repeated until a predefined minimum pattern size is reached to find as many as possible patterns without items relevant to the movement in the area and smallest possible patterns with such items. That is, any of the smaller patterns still covering at least one relevant item can be further divided, i.e., zoomed to yet smaller patterns until the smallest predefined pattern size is reached, thus giving a small area that can be processed with reasonable resources.
If a division of a pattern results smaller patterns where each of the patterns still covers at least one relevant item, it may still be considered advantageous to divide these patterns into next level of smaller patterns. The already divided patterns may still contain some smaller patterns that are free of relevant items, and thus it may be worth dividing to next zoom level, closer to, or until maximum zoom level. A jump point algorithm may then be used for the found tile with at least one relevant item.
In the following the process of dividing the patterns into smaller patterns is called zooming. A purpose of zooming is to reduce the size of pattern(s) covering area(s) with items to be avoided to its minimum where after a path finding algorithm needs to be used to avoid collision with the items only in the determined small pattern(s).
In the example of
The patterns can be called ‘zoom tiles’. The zoom tiles can have zoom levels from 0 to ZMax. A zoom tile of a level (or size) that does not cover any part of any obstacle item, for example NFZ, can be provided with an indication of this in the database of the control system. Such tile does not need any more zoom in for the reason that it is determined to be free of obstacles, i.e., there is a full freedom of movement within the tile. Thus such tile can be indicated as free movement area and there is no need for use of a specific path finding algorithm.
Any tile that does cover at least a part of any movement restricting obstacle item such as a NFZ is marked accordingly in the control system data. Determination of at least one item in the tile triggers zooming of the tile into a next level. As shown in the example of
The zooming operation is repeated until a tile is determined to be free of relevant items, or the defined maximum zoom level ZMax is reached for a tile. A path finding algorithm can then selectively be applied to the tiles with relevant items on the maximum zoom level. Elsewhere it can be assumed that the drone or another autonomous vehicle has the freedom to follow a direct or otherwise desired path through the tiles.
Zoom tiles for a controlled area can be initialized once, e.g. when configuring the system at the time of it being taken into use and/or when maintenance is provided. Update may be needed only if there is a need to apply changes in movement obstacles in the controlled area. For example, a new high structure erected within the area may necessitate update of NFZ data in a GCS.
Zooming can be pre-set in the system and/or provided at runtime. In pre-set zooming the zooming data is defined in a database when the area is configured in the system database and then the already configured tiles are used when determining the flight path. In runtime operation the zooming is provided during path determining operation.
In the
Any appropriate path search algorithm like 2D or 3D path finding search algorithm may be used for path determining within grey tiles that have reached the maximum zoom level. For example, a “Jump Point” algorithm may be used. In the example the shortest, i.e., direct search path 153 extends between locations A in tile 140 and B in tile 142. The search path 153 extends via the white tile 148 and the grey tile 151. As explained above, tiles 148-151 are on the maximum zoom level, the zooming from the previous level (area covered by tile 147) being needed because there is at least one NFZ in the area. Tile 147 comprising the smaller level tiles 148, 149, 150 and 151 is denoted by stronger border line for illustration purposes.
After zooming to the maximum level it can be determined that tile 148 is free of NFZs but tile 151 at the maximum zoom level has at least one NFZ. This triggers use of a path finding algorithm through tile 151 between tile entrance location, denoted by reference An, and tile exit location, denoted by reference Bn, i.e. intersecting parts of tile boundaries of the path. The final path of travel is a combination of the direct parts of the search path 153 through the white tiles 140, 148 and 142 and the product 154 of the path finding algorithm through the grey tile 151. In one example the tiles may be numbered sequentially with a first, second and third value, based on longitude, latitude and zoom level, respectively, illustrating the tile. Location points An and Bn can be calculated from the tile and the path A-B using calculated tile corner coordinates.
In one example embodiment GCS computer may monitor the path planning in relation to its memory usage or any other device having similar functionality. The path is created by adding waypoints between start and target points. If no NFZs are found in the path including start and target points, direct path is ok.
If the tile which comprises one or more NFZs in the desired path, the GCS computer may select different path finding algorithm in the tile where NFZs are found than for the other path having free of the NFZs based on, at least in part, memory needed for path planning.
The tile may have intersection points An and Bn with the path between A and B in
As the smaller the tile is, the more memory is used, the selection can be further influenced by algorithm memory use. It is one option to limit the zooming level by monitoring the memory usage of the GCS computer depending on the algorithm used. There may be a memory usage threshold until the memory usage is not any more useful. The threshold may be different for the used algorithm and/or depend on the zooming level. One or more of the selection criteria might be used in path finding algorithm selection. The jump point search algorithm for the tile, where NFZs exist/s, may be used, for example.
Looping clean-up algorithm is used for cleaning and shortening the paths and/or reducing one or more waypoints received by jump point search algorithm as described further in the application.
In one example embodiment when the predefined minimum zooming level is reached, the tile, which has the NFZs, jump point search algorithm for generating path around the NFZs can be selected for the NFZs, in
In some embodiments the zooming level may be limited to size of tile 143, i.e. coarser tile than the tile 151. In that case the path AB has used its memory, so that it cannot use other coarse grid based path finding algorithm for smaller tiles than size of tile 143.
The path finding algorithm is needed for the area covered by the grey tile 151 since elsewhere there are only white tiles having no NFZ areas on the search path 153.
In operation according to an example, a path search algorithm from start location A to target location B starts with a search by tracing a direct line path from location A to location B on zoom level zero. If a found path does not go over any “grey” tiles it can be determined that there is a direct line A-B with length (A,B). Thus no detailed path finding operation and use of a specific path finding algorithm is needed.
The zoom level can be increased after determination of zoom level Z to zoom level Z+1. A determination can then be made if there is a path length through a particular grey tile on zoom level Z+1. If such grey tile is recognised, a further zoom level is analysed. At the highest resolution zoom level (ZMax) a recognised at least one grey tile can then be subjected to a path finding algorithm after. This may be provided in response to determining by the path search algorithm that the search path goes through at least one grey tile at the zoom level ZMax. The process can repeated for each “grey” tile until an optimal path between locations A and B has been found.
The number of zoom levels can depend on the application. For example, for certain outdoor applications a maximum zoom level ZMax=18 may be considered adequate. For certain indoor applications a maximum zoom level ZMax=21 may be considered adequate. However, these are only examples, and other zoom and considerably different number of zoom levels may also be used.
It is also possible that no path is found over a tile. If no way over a tile is found, any path involving the tile would be blocked as a result. The algorithm can be configured to find a shortest, or otherwise optimal, non-blocked path around the tile from several examined options. A path examination results a path length. If a path is blocked in a tile infinity can be added to the length of the path to select a shortest alternative path. When possible alternative paths have been analysed, simple sorting by length procedure can give a good, or even the best candidate. If infinity setting is used, all blocked path candidates can be placed at the end of the list of candidates. Blocked paths may also be removed from the candidate list.
It is also possible that the path finding algorithm returns to larger tiles, i.e. from analysis of zoom level Z+1 to zoom level Z. This may be desired, e.g., when there is ambiguity or no grey tile is found at the highest resolution zoom level.
The clean-up algorithm may output a path requiring alteration of the entrance and/or exit points. In such case the entire path between start and target locations may need to be altered accordingly.
In one embodiment the pattern needed is identified by zoom level identifier of the pattern and/or pattern coordinates (comparing coordinates of one or more polygon lines with the search path line) and comparison with data in the obstacle database, or bounding box area of an obstacle.
It can then be determined at 174 that a search path between a first location and a second location extends through at least one of the patterns having at least one item relevant to movement. A path finding algorithm can then be used at 175 to determine a path of movement within the determined at least one pattern through which the search path extends.
In accordance with an example the method comprises adjusting at least one of the points where the search path crosses the border of a pattern.
In accordance with an example grid 185 of
The local grid 185 for multiple no-fly zones or other no-go areas can also be provided on xy coordinate system for example such that when there are NFZs 182 and 183 in a local area, the left hand side of the NFZ 182 provides x min value. If this NFZ is also highest/extends furthest in the y direction it also provides y max value to the grid. The right-hand side of NFZ 183 then provides x max value. Again, because NFZ 183 provides the lowest point on the coordinate system it can also provide y min value to the grid in
Both solutions to define the grid can be used, either separately or combined. For example, there can be e.g. one NFZ close to end of the path and two NFZs which are close to the start. The local grid of
In accordance with a possible operation an initial grid is received based on the zooming tiles. In the initial scenario the parameters can include UAV coordinates, target location coordinates, and no-fly zone polygon coordinates. To find a path from UAV location to the target location without crossing the no-fly zone a 2D grid that corresponds to the real-world area may be created. Existing 2D grid-based path finding algorithm is applied to the 2D grid. The solution path is converted to back to real-world coordinates where after the path may be cleaned from excess waypoints.
In accordance with a possible scenario there can be a long distance from start to target location while there are also some no-fly zones scattered around the area. To avoid creating a massive grid and avoid performance issues, grids can be crated only around no-fly zones that are on the direct path from start to target. Each of the grid is then solved as explained above. Overlapping grids may be combined to one grid.
The configuration can be such that the path finding algorithm is only used to patterns with items relevant to movement that have reached a predefined pattern size. A pattern that has reached the maximum zoom level (ZMax), i.e., covers the predefined smallest area, still covers any part of a no-go area such as a NFZ can be handled using any appropriate path finding algorithm. For example, an area within an identified pattern with an NFZ can be prepared for and handled by an appropriate 2-dimensional or 3-dimensional path finding algorithm with required accuracy. The larger zoom patterns can be assumed to be free of no-go areas, and therefore the path can be defined using a straightforward algorithm, for example be given a straight shortest path through the pattern.
A path search algorithm may create several paths and calculate lengths for each of the paths. The algorithm can be configured to return the shortest path as a solution. Other parameters like wind speed, direction of wind may be used as additional parameters when deciding the shortest path.
These principles can be used, for example, for drone path generator over larger operation areas. For example, control areas covering land areas greater than 100×100 km with accuracy of several meters can be provided. It is possible to add and remove non-fly zones or the like areas in run time, since a non-fly zone may affect only one or a few zoom tiles.
In accordance with an aspect apparatuses, methods and computer program code for movement control of a device in a controlled area are arranged to operate without the zoom feature. An apparatus may comprise memory for storing information of patterns and/or grids covering sectors of a controlled area, the stored information comprising determination whether there is at least one item relevant to movement in the area. Processor apparatus can be configured for determining that a search path between a first location and a second location crosses at least one pattern or tile having at least one item relevant to movement in the area for using a path finding algorithm to determine a path of movement within the determined at least one pattern having at least one item relevant to movement in the area.
In accordance with an aspect a method for configuring data in database storage for movement control of a device is provided. The method comprises configuring patterns in the data storage covering sectors of a controlled area, the configuring comprising determining at least one pattern that has at least one item relevant to movement in the area, dividing the determined at least one pattern into smaller patterns, determining at least one of the smaller patterns with at least one item relevant to movement in the area, and repeating the dividing step until predefined smallest pattern size is reached. An appropriate data processing apparatus configured for the computations and/or computer code product can also be provided. Such method and apparatus can be used to configure zoom data in the data storage for a controlled area. In accordance with an aspect a method for movement control of a device comprises obtaining data defining patterns covering sectors of a controlled area, the patterns comprising at least one pattern of a predefined smallest pattern size and having at least one item relevant to movement in the area, the pattern being divided from a larger pattern to the predefined smallest pattern size, determining that a search path between a first location and a second location extends through at least one of the patterns having at least one item relevant to movement, and using a path finding algorithm to determine a path of movement within the determined at least one pattern through which the search path extends. An appropriate data processing apparatus configured for the necessary computations and/or computer code product can also be provided. Such method and apparatus may be used to utilise configured zoom data for a controlled area in planning a route in the area.
In accordance with a non-limiting example control information may be transmitted to an unmanned device as Micro Air Vehicle Link (MAVLink) commands. MAVLink is an open source, point-to-point communication protocol used between a ground control station and unmanned vehicles to carry telemetry and to command and control unmanned vehicles. It may be used to transmit the orientation of an unmanned vehicle, its GPS location and speed. The MAVLink protocol operates at the application layer. It is noted that MAVLink is only given herein as an illustrative example of a protocol operating at this level for this purpose, and other protocols and message sizes may be used instead of this.
Unmanned vehicles may form a swarm. One of such unmanned vehicles may be configured to act as the leader of the swarm, and a path of movement may only need to be defined for the leader.
It is noted that although
In an example causing the apparatus further to determine the first and second points to be points between which grid based path finding algorithm around the obstacle is used. In further example of the example causing the apparatus further determine grid cells provided with the at least two turn points in the determined path around the obstacle, convert the grid cell coordinates of the at least two turn points and the first and second points in the determined path around the obstacle to real-world coordinates, select at least three points from the at least two turn points and the first and second points wherein the three points are consecutive, determine a line between the first point and third point of the consecutive points, determine whether the line and the obstacle intersect, if no intersection, remove the middle point of the consecutive points and use the line as part of the straightened path, or if intersection found keep the second waypoint as part of the straightened path.
The control apparatuses described herein can comprise appropriate circuitry. As used in this specification, the term “circuitry” may refer to one or more or all of the following: (a) hardware-only circuit implementations (such as implementations in only analog and/or digital circuitry) and (b) combinations of hardware circuits and software, such as (as applicable): (i) a combination of analog and/or digital hardware circuit(s) with software/firmware and (ii) any portions of hardware processor(s) with software (including digital signal processor(s)), software, and memory(ies) that work together to cause an apparatus, such as a mobile phone or server, to perform various functions) and (c) hardware circuit(s) and or processor(s), such as a microprocessor(s) or a portion of a microprocessor(s), that requires software (e.g., firmware) for operation, but the software may not be present when it is not needed for operation.” This definition of circuitry applies to all uses of this term in this application, including in any claims. As a further example, as used in this application, the term circuitry also covers an implementation of merely a hardware circuit or processor (or multiple processors) or portion of a hardware circuit or processor and its (or their) accompanying software and/or firmware. The term circuitry also covers, for example and if applicable to the particular claim element, a baseband integrated circuit or processor integrated circuit for a mobile device or a similar integrated circuit in server, a cellular network device, or other computing or network device.
The required data processing apparatus and functions may be provided by means of one or more data processors. The described functions may be provided by separate processors or by an integrated processor. The data processors may be of any type suitable to the local technical environment, and may include one or more of general purpose computers, special purpose computers, microprocessors, digital signal processors (DSPs), application specific integrated circuits (ASystem InformationC), gate level circuits and processors based on multi core processor architecture, as non-limiting examples. The data processing may be distributed across several data processing modules. A data processor may be provided by means of, for example, at least one chip. Appropriate memory capacity can be provided in the relevant devices. The memory or memories may be of any type suitable to the local technical environment and may be implemented using any suitable data storage technology, such as semiconductor based memory devices, magnetic memory devices and systems, optical memory devices and systems, fixed memory and removable memory. One or more of the steps discussed in relation to the flow and signaling charts may be performed by one or more processors in conjunction with one or more memories.
An appropriately adapted computer program code product or products may be used for implementing the embodiments, when loaded or otherwise provided on an appropriate data processing apparatus. The program code product for providing the operation may be stored on, provided and embodied by means of an appropriate carrier medium. An appropriate computer program can be embodied on a computer readable record medium. A possibility is to download the program code product via a data network. In general, the various embodiments may be implemented in hardware or special purpose circuits, software, logic or any combination thereof. Embodiments of the inventions may thus be practiced in various components such as integrated circuit modules. The design of integrated circuits is by and large a highly automated process. Complex and powerful software tools are available for converting a logic level design into a semiconductor circuit design ready to be etched and formed on a semiconductor substrate.
A control apparatus for controlling a device may comprise means for movement control in an area, the apparatus comprising means for determining travel between a first point and a second point by determining an initial path past an obstacle between the first point and the second point, the initial path comprising at least a third point and a fourth point between the first point and the second point, the at least two points between the first point and the second point being turn points where a direction of the travel changes, and in response to determining that a line between two of the points bypassing at least one of the turn points does not intersect the obstacle, removing the at least one turn point from the initial path to determine a straightened path of travel.
It is noted that whilst embodiments have been described in relation to certain architectures, similar principles can be applied to other systems. Therefore, although certain embodiments were described above by way of example with reference to certain exemplifying architectures for wireless networks, technologies standards, and protocols, the herein described features may be applied to any other suitable forms of systems, architectures and devices than those illustrated and described in detail in the above examples. It is also noted that different combinations of different embodiments are possible. It is also noted herein that while the above describes exemplifying embodiments, there are several variations and modifications which may be made to the disclosed solution without departing from the spirit and scope of the present invention.
| Number | Date | Country | Kind |
|---|---|---|---|
| 20195897 | Oct 2019 | FI | national |
