Systems and methods for autonomous movement of materials

Abstract

A method may include: determining, using a controller (136) located on an at least partially autonomously controlled machine (104), a midpoint (204) of a stockpile (118) of material (119) as sensed by a sensor (134); determining, using the controller (136), a plurality of potential loading paths about the midpoint (204) for loading the machine (104); performing, using the controller (136), a cost function analysis of the effectiveness of the potential loading paths based on at least one factor to obtain a most effective loading path; and instructing the machine (104) to perform a loading instance defined by the most effective loading path.

Description

Technical Field

[0001] The present disclosure relates to systems and methods for the movement of material. More particularly, the present disclosure relates to systems and methods for loading an at least partially autonomous machine with material from a stockpile in an efficient manner. Background Art

[0002] In some applications, soil, gravel, minerals, topsoil, and / or other materials can be moved from one location to another on a work site. For example, material can be excavated from an off-site location and moved to the work site to serve as construction material for the work surface at the work site. The material can be temporarily stored in a pile for later distribution to different parts of the work site. A machine, such as a loader, can load a work tool, such as a bucket, with the material. The work tool can be coupled to and actuated by the loader. The machine can access the pile of material, dig into it, and shovel the material into the bucket. Human operators can be trained and gain experience to understand how to most efficiently load the material into the bucket. However, human operators can be expensive to use and may increase potential liability if an accident occurs on the work site. Therefore, autonomous or semi-autonomous machines can be used to improve work site safety and increase the operational efficiency associated with the movement of piles or materials.

[0003] A loader can repeatedly travel from a material pile to a material dumping area until the desired amount of material has been displaced from the pile to the dumping area. A loading entry point can be defined as a point at the edge of the pile where the loader's work tool digs into the pile to obtain a load of material from the pile. The loading direction is the direction of travel of the loader at the entry point and can include a direction perpendicular to or angled with respect to the face of the pile at the entry point. Some example systems may utilize an autonomous or semi-autonomous loader that selects the shortest path between the material pile and the dumping area to determine the loading entry point and / or loading direction for the pile or material. Other example systems may use a direction perpendicular to the edge of the material pile to determine the loading entry point and / or loading direction. Still other example systems may select a direction pointing to a predefined or arbitrary point on the pile or material, or a direction that follows the machine's current orientation. Consequently, the material pile may be undesirably separated into multiple separate piles that remain to be cleaned, and additional loading operations may be required to load the material into the machine's bucket. Furthermore, creating multiple separate piles can spread the material over a larger area of ​​the work site and spread it too thinly along the surface of the work site, resulting in a larger portion of the material being lost because the material may not be captured by the machine.

[0004] Examples of the present disclosure are directed to overcoming the above-mentioned deficiencies. Summary of the Invention

[0005] In an example of the present disclosure, a method may include: sensing an outer surface of a pile of material using a sensor; determining a midpoint of the pile of material based on the sensed outer surface using a controller located on a machine that is at least partially autonomously controlled; determining a plurality of potential loading paths around the midpoint for loading the machine using the controller; selecting a master loading path of the potential loading paths based on a cost function analysis using the controller; and causing the machine to execute a loading instance defined by the master loading path using the controller.

[0006] In another example of the present disclosure, a system includes an at least partially autonomous machine configured to travel along a work surface at a work site. The machine includes a work tool configured to carry material as the machine travels along the work surface. The system also includes a sensor configured to sense an outer surface of a stockpile of material; and a controller in communication with the sensor and the machine. The controller is configured to determine a midpoint of the stockpile of material based on the sensed outer surface; determine a plurality of potential load paths for loading the machine about the midpoint; and cause the machine to execute a load instance defined by the primary load path.

[0007] In yet another example of the present disclosure, a system may include: an at least partially autonomously controlled machine including a work tool configured to carry material as the machine travels along a work surface; a sensor configured to sense an outer surface of a stockpile of material and determine a midpoint of the stockpile of material; and a controller in communication with the sensor and the machine via a communication network. The controller is configured to determine the midpoint of the stockpile of material based on the sensed outer surface, determine a plurality of potential load paths around the midpoint for loading the machine, select a primary load path of the potential load paths based on a cost function analysis, and cause the machine to execute a load instance defined by the primary load path. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 is a schematic diagram of a system according to an example embodiment of the present disclosure.

[0009] Figure 2 yes Figure 1 Another schematic diagram of the system shown in .

[0010] Figure 3 It is depicted with Figure 1 and Figure 2 Flowchart of an example method associated with the system shown in .

[0011] Figure 4 It is depicted with Figure 1 and Figure 2Flowchart of another example method associated with the system shown in . DETAILED DESCRIPTION

[0012] Wherever possible, the same reference numbers will be used throughout the drawings to refer to the same or like parts. Figure 1 is a schematic diagram of a system 100 according to an example embodiment of the present disclosure. Figure 1 The example system 100 shown in FIG. 1 may include one or more machines operating at a work site 112 to perform various tasks. For example, the machines may include one or more excavators 102, one or more loaders 104, one or more haulers 106, such as haul trucks, and / or other types of machines, such as pavers, used for construction, mining, paving, excavation, and / or other operations at the work site 112. Each of the machines described herein may communicate with each other and / or with a local or remote control system 120 via one or more central stations 108. The central station 108 may facilitate wireless communication between the machines described herein and / or between such machines and, for example, a system controller 122 of the control system 120, for the purpose of transmitting and / or receiving operational data and / or instructions.

[0013] An excavator 102 may refer to any machine that reduces material at a work site 112 for the purpose of subsequent operations (i.e., for blasting, loading, hauling, and / or other operations). Examples of excavators 102 may include excavators, backhoes, bulldozers, drills, trenchers, and tow ropes, among others. Multiple excavators 102 may be co-located within a common area at a work site 112 and may perform similar functions. For example, one or more of the excavators may move soil, sand, minerals, gravel, concrete, asphalt, topsoil, and / or other materials comprising at least a portion of a work surface 110 at a work site 112. Thus, under normal conditions, each of the co-located excavators 102 may have similar productivity and efficiency when exposed to similar site conditions.

[0014] Loader 104 may refer to any machine that lifts, carries, loads, and / or removes material that has been removed by one or more of excavators 102. In some examples, loader 104 may remove such material and may transport the removed material from a first location at worksite 112 to a second location at worksite 112. Examples of loader 104 may include a wheel loader or track loader, a front shovel, an excavator, a cable shovel, a stacker-reclaimer, or any other similar machine. One or more loaders 104 may operate within a common area of ​​worksite 112, for example, to load the removed material onto hauler 106. For example, the loaders 104 described herein may traverse one or more travel paths across work surface 110, as defined by a method for identifying effective ways to move material 119 and effective and efficient travel paths. In one example, such travel paths may include those between a stockpile 118 of material 119 and a dump site for the material, which may prove most efficient in terms of time, distance, and fuel usage by the loaders 104. In another example, such a travel path may include one or more partially or fully formed roads, bridges, tracks, paths, or other surfaces formed by the work surface 110 and traversable by the construction, mining, paving, and / or other example machines described herein. In this example, one such travel path may extend from a pile 118 or a collection of other materials 119 removed by one or more excavators 102 and / or loaders 104 at the work site 112 to numerous other locations within the work site 112. A detailed description of the paths that the excavators 102 and / or loaders 104 may take when moving material is presented in greater detail herein. Furthermore, the manner in which the excavators 102 and / or loaders 104 may process the pile 118 of material 119 and how the excavators 102 and / or loaders 104 may use their respective work tools 140 to dig into the pile and obtain a full load, or at least a full load, as effectively and efficiently as possible is also described below. In some examples, one or more ditches, ruts, potholes, accumulations or stockpiles 118 of material 119 or other defects may be provided on or formed by the working surface 110. In some examples, and as Figure 1As shown, a pile 118 can be located along one or more travel paths of a loader 104 or other machine described herein. In such an example, the system 100 can be configured to identify the pile 118 and determine various travel parameters for the machine (e.g., alternative travel paths, travel speed of the machine, etc.) based at least in part on identifying the pile 118. Controlling the machine to operate based on such travel parameters can reduce the time and resources required for a machine (e.g., loader 104) to complete a desired task, can reduce the risk of damage to the machine, and can improve the overall efficiency of the system 100. Controlling the machine to operate based on such travel parameters can also reduce the risk of harm or injury to the operator of the machine. Hauler 106 can refer to any machine that carries excavated material between different locations within the worksite 112. Examples of haulers 106 can include articulated trucks, off-highway trucks, on-highway dump trucks, wheeled tractor-scrapers, or any other similar machine. The loaded hauler 106 can carry topsoil from the excavation area within the work site 112 along haul roads to various dumps and return to the same or a different excavation area for another load. Figure 1 In some examples, the control system 120 and / or the system controller 122 may be located at a command center (not shown) remote from the work site 112. In other examples, the system controller 122 and / or one or more components of the control system 120 may be located at the work site 112. Regardless of the location of the various components of the control system 120, such components may be configured to facilitate communication between and provide information to the excavators 102, loaders 104, haulers 106, and / or other machines of the system 100 (hereinafter collectively referred to as equipment). In any of the examples described herein, the functionality of the system controller 122 may be distributed such that certain operations are performed at the work site 112 and other operations are performed remotely (e.g., at the remote command center described above). For example, some operations of the system controller 122 may be performed at the work site 112 on one or more excavators 102, one or more loaders 104, one or more haulers 106, and / or the like. It should be understood that system controller 122 may include components of system 100 , components of one or more machines located at worksite 112 , components of a separate mobile device (eg, a mobile phone, tablet computer, laptop computer, etc.), and / or control system 120 .

[0015] The system controller 122 may be an electronic controller that operates in a logical manner to perform operations, run control algorithms, store and retrieve data, and perform other required operations. The system controller 122 may include or have access to memory, auxiliary storage devices, a processor, and any other components used to run / execute application programs. The memory and auxiliary storage devices may be in the form of read-only memory (ROM) or random access memory (RAM), or integrated circuits accessible by the controller. Various other circuits may be associated with the system controller 122, such as power supply circuitry, signal conditioning circuitry, driver circuitry, and other types of circuitry.

[0016] The system controller 122 may be a single controller or may include more than one controller (e.g., an additional controller associated with each device) configured to control various functions and / or features of the system 100. As used herein, the term "controller" is used broadly to encompass one or more controllers, processors, central processing units, and / or microprocessors that may be associated with the system 100 and that may cooperate to control various functions and operations of the machines included in the system 100. The functionality of the system controller 122 may be implemented in hardware and / or software, regardless of function. The system controller 122 may rely on one or more data graphs, lookup tables, neural networks, algorithms, machine learning algorithms, and / or other components related to the operating conditions and operating environment of the system 100, which may be stored in the memory of the system controller 122. Each of the aforementioned data graphs may include a collection of data in the form of tables, graphs, and / or equations to maximize the performance and efficiency of the system 100 and its operations.

[0017] Components of control system 120 can communicate with and / or be otherwise operatively connected to any component of system 100 via network 124. Network 124 can be a local area network ("LAN"), a larger network, such as a wide area network ("WAN"), or a collection of networks, such as the Internet. A protocol for network communications, such as TCP / IP, can be used to implement network 124. Although embodiments are described herein as using a network 124 such as the Internet, other distribution techniques for transmitting information via memory cards, flash memory, or other portable storage devices can be implemented.

[0018] It should also be understood that the device may include a respective controller, and each of the respective controllers described herein (including the system controller 122) may communicate and / or be otherwise operably connected via a network 124. For example, the network 124 may include components of the wireless communication system of the system 100, and as part of such a wireless communication system, the device may include a respective communication device 126. Such a communication device 126 may be configured to allow for wireless transmission of a plurality of signals, instructions, and / or information between the system controller 122 and the respective controller of the device. Such a communication device 126 may also be configured to allow for communication with other machines and systems remote from the worksite 112. For example, such a communication device 126 may include a transmitter configured to transmit signals (e.g., via the central station 108 and over the network 124) to receivers of one or more other such communication devices 126. In such instances, each communication device 126 may also include a receiver configured to receive such signals (e.g., via the central station 108 and over the network 124). In some examples, the transmitter and receiver of a particular communication device 126 may be combined into a transceiver or other such component. In any of the examples described herein, such a communication device 126 may also enable communication with one or more tablet computers, computers, cellular / wireless phones, personal digital assistants, mobile devices, or other electronic devices 128 located at and / or remote from the worksite 112 (e.g., via the central station 108 and over the network 124). Such electronic devices 128 may include, for example, a mobile phone and / or tablet computer of a project manager (e.g., a foreman) who oversees daily operations at the worksite 112.

[0019] The network 124, communication device 126 and / or other components of the wireless communication system described above can implement or utilize any desired system or protocol including any of a plurality of communication standards. The desired protocol will allow communication between the system controller 122, one or more of the communication devices 126 and / or any other desired machines or components of the system 100. Examples of wireless communication systems or protocols that can be used by the system 100 described herein include wireless personal area networks (e.g., IEEE 802.15) such as Bluetooth RTM, local area networks such as IEEE 802.11b or 802.11g, cellular networks, or any other systems or protocols for data transmission. Other wireless communication systems and configurations are contemplated. In some cases, wireless communication can be sent and received directly between the control system 120 and the machines (e.g., paving machines, haul trucks, etc.) of the system 100 or between such machines. In other cases, communication can be automatically routed without the need for remote personnel to retransmit.

[0020] In an example embodiment, one or more machines of system 100 (e.g., one or more excavators 102, loaders 104, haulers 106, etc.) may include position sensors 130 configured to determine the position, speed, heading, and / or orientation of the respective machines. In such embodiments, the communication device 126 of the respective machines may be configured to generate and / or transmit signals indicating such determined position, speed, heading, and / or orientation to, for example, system controller 122 and / or other respective machines of system 100. Furthermore, in one example, the position sensors 130 of each of the machines may be configured to assist the machine in determining its position relative to, for example, a pile 118 of material 119. In this example, the position sensors 130 may be used to determine an efficient path to the pile 118 of material 119 and a plurality of entry points when loading a portion of the pile 118 of material 119 into the machine's work tool 140. Furthermore, in this example, the position sensors 130 may also be used to determine the number of loading directions for each of the plurality of entry points. More is described herein regarding the function of position sensor 130 in loading stockpile 118 of material 119 into a machine.

[0021] In some examples, the location sensor 130 of the respective machine may include and / or contain components of a global navigation satellite system (GNSS) or a global positioning system (GPS). Alternatively, a universal total station (UTS) may be utilized to locate the respective location of the machine. In some embodiments, one or more of the location sensors 130 described herein may include a GPS receiver, transmitter, transceiver, laser prism, and / or other such devices, and the location sensor 130 may communicate continuously, substantially continuously, or at various time intervals with one or more GPS satellites 132 and / or UTS to determine the respective location of the machine to which the location sensor 130 is connected. One or more additional machines of the system 100 may also communicate with one or more GPS satellites 132 and / or UTS, and such GPS satellites 132 and / or UTS may also be configured to determine the respective location of such additional machines. In any of the examples described herein, the system controller 122 and / or other components of the system 100 may use the machine position, speed, heading, orientation, and / or other parameters determined by the respective position sensors 130 to coordinate activities of the excavator 102, the loader 104, the hauler 106, and / or other components of the system 100. Furthermore, in any of the examples described herein, the system controller 122 and / or other components of the system 100 may use the machine position, speed, heading, orientation, and / or other parameters determined by the respective position sensors 130 to move portions of the pile 118 of material 119 throughout the worksite 112 as described herein.

[0022] In one example, the position determined by the position sensor 130 carried by the loader 104 can be used by the system controller 122 and / or the controller 136 of the loader 104 to determine a travel path extending from the current position of the loader 104 to the location of the stockpile 118 of material 119 and to the location of a dump within the work site 112. In such an example, the controller 136 of the loader 104 can control the loader 104 to traverse at least a portion of the work site 112 along the path to complete one or more tasks at the work site 112, including moving material within the stockpile 118. In addition, the controller 136 of the loader 104 can control the loader 104 to approach the stockpile 118 of material 119 along an efficient path. Further, the controller 136 of the loader 104 can control the loader 104 to excavate from a plurality of entry points using at least one of a plurality of loading directions. This determined travel path may be used to maximize the operating efficiency of the loader 104 in moving material from the pile 118 to other portions of the worksite 112, and generally maximize the efficiency of the system 100. In the examples described herein, the controller 136 may be or may include an electronic control module (ECM).

[0023] In any of the examples described herein, the system controller 122 and / or the respective controllers 136 of the various machines of the system 100 can be configured to generate a user interface displayed on a display device associated with the system controller 122 and / or the machines 102, 104, 106, the user interface including information indicating the travel path, travel speed, orientation, and / or other travel parameters of the respective machines. Furthermore, the system controller 122 and / or the respective controllers 136 can be configured to generate information on the user interface indicating the path, number of entry points, and / or number of loading directions used by the loader 104 to access the stockpile 118 of material 119.

[0024] In some examples, and in addition to the various travel parameters described above, the system controller 122 and / or the controller 136 of the loader 104 may also determine one or more work tool positions associated with the work tool 140 of the loader 104. In such examples, the user interface may also include information indicating the determined work tool positions. As described in greater detail below, each work tool position may correspond to a respective position along at least one of the travel paths, a path used by the loader 104 to access the stockpile 118 of material 119, a number of entry points, and / or a number of loading directions. In any of the examples described herein, such a user interface may be generated by the controller 136 and provided to, for example, the electronic device 128 (e.g., via the network 124), a display of the loader 104, the system controller 122 (e.g., via the network 124), and / or one or more components of the system 100 for display. Additionally or alternatively, such a user interface can be generated by the system controller 122 and provided to, for example, the electronic device 128 (e.g., via the network 124), a display of the loader 104, the controller 136 (e.g., via the network 124), and / or one or more components of the system 100 for display. In any of the examples described herein, one or more devices can be manually controlled, semi-autonomously controlled, and / or fully autonomously controlled. In examples where devices are operated under autonomous or semi-autonomous control, the speed, steering, work tool positioning / movement, and / or other functions of such machines can be automatically or semi-autonomously controlled based at least in part on the determined travel parameters and / or work tool position described herein.

[0025] Continue to refer Figure 1 As described above, each device may include a controller 136, and such controller 136 may include components of a local control system onboard and / or otherwise carried by the respective machine. Such controllers 136 may be generally similar to or identical to the system controller 122 of the control system 120. For example, each such controller 136 may include one or more processors, memory, and / or other components described herein with respect to the system controller 122. In some examples, the controller 136 may be located on a respective one of the loaders 104 and may also include components located remotely from the respective one of the loaders 104, such as on any other machine in the system 100 or at a command center as described herein. Thus, in some examples, the functionality of the controller 136 may be distributed, such that certain functions are performed on the respective one of the loaders 104 and other functions are performed remotely. In some examples, the controller 136 of the local control system carried by the respective machine may, alone or in combination with the control system 120, implement autonomous and / or semi-autonomous control of the respective machine.

[0026] Furthermore, in addition to the communication device 126 and position sensor 130 described above, one or more of the devices may include a perception sensor 134 configured to determine one or more characteristics of the work surface 110. For example, a controller 136 of a particular machine may be electrically connected to and / or otherwise in communication with the communication device 126, the position sensor 130, and the perception sensor 134 carried by the particular machine, and the perception sensor 134 may be configured to sense, detect, observe, and / or otherwise determine various characteristics of the work surface 110. In one example, the perception sensor 134 may be configured to sense, detect, observe, and / or otherwise determine various characteristics of the stockpile 118 of material 119 to be distributed throughout the worksite 112. The perception sensor 134 may include at least one of a position sensor, an imaging device, or other sensing device. Furthermore, in one example, the GPS satellite 132 may include an imaging device that may capture an image of the stockpile 118 of material 119 and transmit the image to one or more of the devices for use in determining the shape or contour of the stockpile 118 of material 119. Further details regarding the perception sensor 134 are described below.

[0027] In some examples, one or more of communication device 126, position sensor 130, and perception sensor 134 may be fixed to a cab, chassis, frame, and / or any other component of a respective machine. However, in other examples, one or more of communication device 126, position sensor 130, and perception sensor 134 may be removably attached to a respective machine and / or disposed, for example, within a cab of such a machine during operation.

[0028] In some examples, perception sensor 134 may comprise a single sensor and / or other components of a local perception system located on a machine (e.g., located on loader 104). In other examples, perception sensor 134 may comprise multiple, identical or different sensors (e.g., a sensor array), each of which comprises a component of such a local perception system located on the machine. For example, perception sensor 134 may comprise an image capture device, among others. Such an image capture device may be any type of device configured to capture images representing work surface 110, worksite 112, pile 118 of material 119, and / or other environments within the image capture device's field of view. For example, example image capture devices may comprise one or more cameras, such as an RGB camera, a monochrome camera, an intensity (grayscale) camera, an infrared camera, an ultraviolet camera, a depth camera, a stereo camera, and the like. Such an image capture device may be configured to capture image data representing, for example, the length, width, height, depth, volume, color, texture, composition, radiation emission, and / or other characteristics of one or more objects, including, for example, pile 118 of material 119 within the image capture device's field of view. For example, such characteristics may also include one or more of x-position (global position coordinates), y-position (global position coordinates), z-position (global position coordinates), orientation (e.g., roll, pitch, yaw), object type (e.g., classification), object velocity, and object acceleration. It should be understood that one or more of such characteristics (e.g., position, size, volume, etc.) may be determined by an image capture device alone or in combination with position sensor 130 as described above. Characteristics associated with work surface 110 and / or worksite 112 may also include, but are not limited to, the presence of another machine, person, or other object in the field of view of perception sensor 134, the time of day, day of the week, season, weather conditions, and indications of darkness / light. In one example, characteristics associated with work surface 110 and / or worksite 112 may be captured in real time by position sensor 130 and / or perception sensor 134. In this example, capturing data in real time from position sensor 130 and / or perception sensor 134 provides advantages for repeatedly moving material 119 from pile 118. For example, as the loader 104 moves between the pile 118 and the dump, the position sensors 130 and / or the perception sensors 134 may capture data defining the position and direction of travel of the loader 104, the position of the pile 118, the shape of the pile, and other characteristics of the work surface 110 and / or worksite 112. This allows the system controller 122 and / or the controller 136 of the loader 104 to more effectively and quickly determine the most efficient travel path for the loader, as well as the entry point and loading direction of the travel path.

[0029] The image capture device and / or other components of the perception sensor 134 may also be configured to provide one or more signals (including such image data or other sensor information captured thereby) to the controller 136 of the machine 102, 104, 106. Such sensor information may include, for example, a plurality of images captured by the image capture device and indicating various characteristics of one or more objects within the image capture device's field of view (including, for example, a stockpile 118 of material 119). Each such image may include a respective set of images of the stockpile 118 and / or other objects detectable by the image capture device. In such instances, the controller 136 and / or the system controller 122 may analyze the sensor information received from the perception sensor 134 to identify the stockpile 118 indicated by the sensor information (e.g., shown or otherwise included in such an image).

[0030] The perception sensors 134 and / or the local perception systems carried by the machines may also include light detection and ranging (LIDAR) sensors. Such LIDAR sensors may include one or more lasers or other light emitters carried by (e.g., mounted on, connected to, etc.) a particular machine 102, 104, 106, and one or more light sensors configured to receive radiation, reflections, and / or radiation returned by an object upon which light from such light emitters has been incident. In an example embodiment, such LIDAR sensors may be configured such that the one or more lasers or other light emitters are mounted to rotate (e.g., about a substantially vertical axis), thereby sweeping the light emitters through a range of motion, for example, 360 degrees, to capture LIDAR sensor data associated with the stockpile 118 of material 119, the work surface 110, and / or the worksite 112. For example, a LIDAR sensor of the present disclosure may have a light emitter and a light sensor, wherein the light emitter includes one or more lasers that direct highly focused light toward an object or surface, which reflects the light back to the light sensor, although any other light emission and detection for determining range is contemplated (e.g., flash LIDAR, MEMS LIDAR, solid-state LIDAR, etc.). Measurements from such a LIDAR sensor may be represented as three-dimensional LIDAR sensor data having coordinates (e.g., Cartesian coordinates, polar coordinates, etc.) corresponding to a location or distance captured by the LIDAR sensor. For example, the three-dimensional LIDAR sensor data and / or other sensor information received from the LIDAR sensor may include a three-dimensional map or point cloud, which may be represented as multiple vectors emanating from the light emitter and terminating at an object (e.g., pile 118) or a surface (e.g., work surface 110). In some examples, a conversion operation may be used by the controller 136 and / or by the system controller 122 to convert the three-dimensional LIDAR sensor data into multi-channel two-dimensional data. In some examples, LIDAR sensor data and / or other sensor information received from perception sensors 134 may be automatically segmented by controller 136 and / or by system controller 122, and the segmented LIDAR sensor data may be used, for example, as input to determining a trajectory, a travel path, a loading entry point, a loading direction, a travel speed, and / or other travel parameters of a machine described herein (e.g., travel parameters of one or more of loaders 104).

[0031] The perception sensors 134 and / or the local perception system carried by the machine may also include one or more additional sensors. Such additional sensors may include, for example, radio detection and ranging (hereinafter referred to as "RADAR") sensors, sound navigation and ranging (hereinafter referred to as "SONAR") sensors, depth sensing cameras, ground-penetrating radar sensors, magnetic field emitters / detectors, and / or other sensors disposed on the vehicle and configured to detect objects present in the work site 112. Each of the sensors described herein with respect to the perception sensors 134 and / or the local perception system may output one or more corresponding signals to the controller 136 and / or the system controller 122, and such signals may include any of the aforementioned sensor information (e.g., image data, LIDAR data, RADAR data, SONAR data, GPS data, etc.). The various sensors of the perception sensors 134 may capture such sensor information simultaneously, and in some cases, the sensor information received from the corresponding sensors of the perception sensors 134 may include identification and / or indication of one or more of the same objects (e.g., the pile 118) sensed by such sensors. In such instances, the controller 136 and / or the system controller 122 may analyze the sensor information received from each respective sensor to identify and / or classify one or more objects indicated by the sensor information.

[0032] For example, the controller 136 and / or the system controller 122 may associate the output of each sensor modality with a specific object and / or a specific location of the worksite 112 stored in its memory. Utilizing such data association, object recognition, and / or object characterization techniques, the output of each of the sensors described herein may be compared. Through such comparisons, and based at least in part on sensor information received from the perception sensors 134 and / or the position sensors 130, the controller 136 and / or the system controller 122 may identify one or more objects located at the worksite 112 (e.g., a pile 118 located within the work surface 110). As described above, the corresponding sensor information received from both the perception sensors 134 and the position sensors 130 may be combined and / or considered together by the controller 136 and / or the system controller 122 to determine the location, shape, size, volume, and / or other characteristics of the pile 118 of material 119 described herein.

[0033] Furthermore, in some examples, and depending on the accuracy and / or fidelity of the sensor information received from the various sensors associated with perception sensors 134, the presence, location, orientation, identity, length, width, height, depth, and / or other characteristics of a pile 118 identified by controller 136 using first sensor information (e.g., LIDAR data) may be verified by controller 136 using second sensor information (e.g., image data) obtained concurrently with the first sensor information but from a different sensor or modality of perception sensors 134. In one example, sensor data obtained from multiple perception sensors 134 located in multiple devices may be shared between the machines and used to provide a more accurate and / or complete picture of a pile 118.

[0034] Continue to refer Figure 1In some examples, one or more machines within system 100, including loader 104, may include an implement or other work tool 140 coupled to the machine's frame. For example, in the case of loader 104, work tool 140 may include a bucket configured to carry material within an open volume or substantially open space. Loader 104 may be configured to scoop, lift, and / or otherwise load material (e.g., material 119 within pile 118) into work tool 140, for example, by lowering work tool 140 to a loading position. For example, loader 104 may include one or more linkages 142 movably coupled to the frame of loader 104. Work tool 140 may be coupled to such linkages 142, and linkages 142 may be used to lower work tool 140 using, for example, one or more hydraulic cylinders, electric motors, or other devices coupled thereto. In this manner, work tool 140 can be lowered to a loading position, wherein a front edge 144 of work tool 140 is positioned proximate to, adjacent to, and / or at work surface 110, and a base of work tool 140 is positioned substantially parallel to work surface 110. Loader 104 can then be controlled to advance toward pile 118 of material 119 so that work tool 140 can impact pile 118 of material 119, positive volumetric defects, and / or other objects positioned on work surface 110. The advancement of loader 104 causes the material to be at least partially transferred into the open space of work tool 140. Linkage 142 can then be controlled to raise, pivot, and / or tilt work tool 140 to a carrying position above work surface 110 and substantially out of the sight of, for example, an operator controlling the movement of loader 104. Loader 104 may then be controlled to traverse work site 112 until loader 104 reaches a dump area, hauler 106, and / or another location at work site 112 designated for receiving the moved material carried by work tool 140. Linkage 142 may then be controlled to lower, pivot, and / or tilt work tool 140 to a dump position in which the material carried within the open space of work tool 140 may be deposited (e.g., due to gravity acting on the material carried by work tool 140) at a dump area, in the bucket of hauler 106, and / or elsewhere as desired.

[0035] Figure 2 yes Figure 1 Another schematic diagram of the system 100 is shown in FIG. Figure 2As shown, in some examples, the work site may include a dump 250 spaced apart from the stockpile 118 of material 119 described above. The loader 104 is depicted traversing a travel path 270 between the stockpile 118 (serving as a loading dock), a back-up point 255, and the dump 250. The back-up point 255 may be any point at which the loader 104 can perform a turn when facing away from the stockpile 118. The loader 104 may use any number of back-up points 255 when moving material 119 from the stockpile 118 to the dump 250 and when moving from the dump 250 to the stockpile 118. Furthermore, the work site 112 may include any number of dumps 250 where material 119 may be deposited.

[0036] In some exemplary operations, the goal of the autonomously or semi-autonomously operated loader 104 may be to move material 119 from the pile 118 to another portion of the work site 112, including the dump 250. In the examples described herein, the loader 104 may obtain information from various sensors to move the material 119 within the pile 118. The information sensed by the perception sensors 134, the position sensors 130, and / or other sensors may be transmitted to the controller 136 of the loader 104 and / or to the system controller 122 via the communication device 126, the central station 108, the satellite 132, the network 124, and / or other communication devices for processing by the controller 136 of the loader 104 and / or the system controller 122. In one example, the information sensed by the perception sensors 134, the position sensors 130, and / or other sensors may be transmitted to the controller 136 of the loader 104, where the controller 136 of the loader 104 processes the information without assistance from resources of the system controller 122. In another example, information sensed by the perception sensors 134, the position sensors 130, and / or other sensors may be transmitted to the system controller 122, where the system controller 122 of the loader 104 processes the information without assistance from resources of the controller 136 of the loader 104. In yet another example, both the system controller 122 and the controller 136 of the loader 104 may participate in the processing of the information.

[0037] and Figure 2 The operations described in Figure 1The system 100 can be used to determine the shape of a pile 118 of material 119. The angle of repose of the material 119 within the pile 118 plays a role in the shape of the pile 118 of material 119 and can be defined as the steepest angle of descent relative to a horizontal plane at which the material 119 can be accumulated without collapsing or sliding (i.e., due to gravity). When granular material is poured onto a horizontal surface, such as the work surface 110 of the work site 112, a generally conical pile is formed. Within the definition of the angle of repose, the internal angle between the surface of the pile and the horizontal surface is related to the density, surface area and shape of the particles, as well as the coefficient of friction of the material. Therefore, based on the composition of the material, such as soil, sand, gravel, etc., the angle of repose can vary.

[0038] When stockpile 118 of material 119 is generally uniform in composition and formed from a single dump of material 119, stockpile 118 may be generally conical in shape. However, when multiple loads of material 119 are dumped together into stockpile 118, multiple conical shapes may form from the loads and blend together into an asymmetrical shape, such as the shape described by initial perimeter 202 of stockpile 118. The shape of stockpile 118 formed in this manner may be described as "potato-shaped" or irregular. Despite its irregular nature, the shape of stockpile 118 obtained from sensors 130, 134 is still useful. For example, perception sensor 134, working in conjunction with position sensor 130, can detect the shape of stockpile 118 as described herein and transmit this information to system controller 122 and / or controller 136 of loader 104 for processing. System controller 122 and / or controller 136 may define the shape of stockpile 118 based on the information. In one example, data obtained from sensors 130, 134 may be used to detect points located along the surface of stockpile 118. Sensors 130, 134 may transmit this data representing the points along the surface of stockpile 118 to, for example, controller 136 of loader 104 and / or system controller 122. Controller 136 of loader 104 and / or system controller 122 may then use this data to create a three-dimensional (3D) point cloud or other structure defined by the points located on the exterior surface of stockpile 118. In one example, controller 136 of loader 104 and / or system controller 122 may apply an image stitching process, a 3D reconstruction process, or other process that combines multiple images or points to generate a 3D image of the exterior surface of stockpile 118.

[0039] In order to move material 119 as efficiently as possible using the determined shape of pile 118, system controller 122 and / or controller 136 of loader 104 may use any information sensed by perception sensors 134, position sensors 130, and / or other sensors described herein to determine a midpoint 204 of pile 118 of material 119. Thus, in any example, the information sensed by perception sensors 134, position sensors 130, and / or other sensors may include information identifying midpoint 204. Midpoint 204 of pile 118 of material 119 may be obtained by analyzing three-dimensional or two-dimensional data obtained by at least one sensor 130, 134 and determining the midpoint of the bulk (i.e., pile 118). Midpoint 204 of pile 118 may be substantially uniform in composition, and the center of mass (i.e., midpoint 204) may be the center of mass of pile 118. In this example, the center of mass may be defined as follows:

[0040]

[0041] where R is the coordinate of the center of mass, M is the total mass of the particles (e.g., soil particles, rocks, gravel, etc.), and m i is the mass of each particle, and r i In this example, the mass of the entire stockpile 118 can be determined by information obtained from a scale or other source, which can provide the mass of the material 119 within the stockpile 118 .

[0042] In another example, midpoint 204 may be determined by determining the center of mass of pile 118 in a single plane or n-dimensional space. In an example where the center of mass is determined in a single plane, for example, the average of the positions of all points within initial perimeter 202 of pile 118 along a single horizontal plane may serve as the single plane, and the center of mass of the shape of the plane may serve as midpoint 204. This example may be most effective in determining midpoint 204 when sensors 130 and 134 have detected all sides of pile 118.

[0043] In an example, when determining the center of mass in n-dimensional space (i.e., in all planes of pile 118), points along multiple planes can be used to identify where mass is located within pile 118. In this example, midpoint 204 can be the average of all points within pile 118 weighted by the local density of material 119. Assuming that pile 118 of material 119 has a substantially uniform density, midpoint 204 is the same as the center of mass of the three-dimensional shape of pile 118.

[0044] In instances where the entire two-dimensional pile shape is known (ie, via data obtained from satellites 132 , position sensors 130 , perception sensors 134 , or a combination thereof), system controller 122 and / or controller 136 of loader 104 may utilize mathematical interpolation to find midpoint 204 .

[0045] In an example of identifying a boundary of the pile facing the loader 104, the system controller 122 and / or the controller 136 of the loader 104 may obtain information sensed by the position sensor 130, the perception sensor 134, or a combination thereof, where the information may be collected from the front side 281 of the pile 118, as opposed to the rear side 280. In this example, the system controller 122 and / or the controller 136 of the loader 104 may apply a polynomial interpolation technique to construct a smooth convex curve connecting the two outermost ends 282-1, 282-2 of the known boundary curve of the front side 281 of the pile 118. The midpoint 204 may then be determined based on the results generated via the polynomial interpolation.

[0046] In some instances and / or situations, if the pile 118 is located on the periphery of the work site 112 and the dump 250 is located on the front side 281, the sensors 130, 134 coupled to the loader 104 may not be able to detect the rear side 280 of the pile 118. Furthermore, driving the loader 104 around the rear side 280 of the pile may be inefficient because it may reduce the operating efficiency of the loader 104, in addition to resulting in wasted time, fuel, and other resources. Therefore, in some examples described herein, the loader 104 may operate without sensing data representing the rear side 280 of the pile 118. In these examples, the loader 104 can operate more efficiently by being instructed to utilize the most efficient loading operation based on the identified and selected entry points 210-1, 210-2, 210-3, 210-4 (collectively referred to herein as 210) and the loading directions 212-1, 212-2, 212-3, 212-4, 212-5, 212-6, 212-7, 212-8, 212-9, 212-10, 212-11, 212-12 (collectively referred to herein as 212) of the entry points 210. For semi-autonomous or autonomous loaders 104, in order to maximize overall efficiency and minimize the amount of leftover material 119 to be picked up and moved via manual and / or remotely controlled machine cleaning devices, the system 100 can select an optimal loading entry point 210 and associated loading direction 212 strategy that takes into account the shape of the stockpile 118 of the material 119.

[0047] Without the present system and method, semi-autonomous and autonomous loaders 104 may rely on extremely primitive approaches to efficiently move the pile 118 of material 119. For example, the shortest path may be selected as the loading entry point. In another example, a direction perpendicular to the edge of the pile 118, a direction toward some predefined point, or a direction that follows the current orientation of the loader 104 may be selected. However, these solutions are generally inefficient. Using these primitive approaches may produce undesirable results, as shown by dashed curves 250-1 and 250-2, where a relatively large pile 118 is divided into multiple, separate, smaller piles that still require eventual removal. Furthermore, using these primitive approaches may produce an undesirable result, as shown by dashed curve 251, where thin, circular layers of material still require eventual removal. The hatching between dashed curves 250-1, 250-2, and 251 and the solid outer line indicating the initial perimeter 202 of the pile 118 indicates that, using these primitive approaches, material 119 may still remain after a significant portion of the pile 118 has been moved. The resulting undesirable shape of the stockpile 118 is depicted by curves 250-1, 250-2, 251. Loading material 119 from the undesirably shaped stockpile 118 may become more difficult and significantly inefficient due to the lack of resistance provided by the convex stockpile 118 of material 119. Obtaining and / or maintaining the convex stockpile 118 during the loading operation of the loader 104 may provide the greatest resistance against the loader 104, which may facilitate the entry of the greatest amount of material 119 into the work tool 140 of the loader 104. Furthermore, using the original method described above, such a loader 104 may also accomplish loading instances in which only a partial load of material 119 may be obtained because there may not be enough material 119 in a significantly sufficient aggregate to constitute a full load of the work tool 140. A user of the loader 104 may desire to have the semi-autonomous or autonomous loader 104 function similarly to the manner in which a human operator would operate a loader. A human operator, having experience with the shape of the stockpile 118 considering the various materials 119, may effectively shovel material 119 from the stockpile 118 in an orderly manner such that, as Figure 2 2. As depicted in FIG, material indicated by curve 220-1 is obtained in a first phase, and material indicated by dashed curve 220-2 is subsequently obtained in a second phase. In this manner, the volume of material 119 within the work tool 140 of the loader 104 can be maximized for each scooping attempt of material 119 while minimizing the energy expended in the eventual cleanup of any loose or unaggregated material 119.

[0048] Thus, after determining the midpoint 204 described herein, the system controller 122 and / or the controller 136 of the loader 104 may identify a plurality of candidate entry points 210 located on the edge of the pile 118 of material 119. In one example, the significant candidate entry points 210 may exclude any entry points 210 located on the back side 280 of the pile 118 of material 119. Furthermore, the significant candidate entry points 210 may exclude any entry points 210 that are further away from the current position of the loader 104 relative to the other significant candidate entry points 210. In this example, those entry points 210 that are further away from the current position of the loader 104 relative to the other significant candidate entry points 210 may be ignored. Furthermore, entry points 210 are significant when the loader 104 and its work tool 140 are moving into the pile 118, rather than out of or away from it.

[0049] Furthermore, in one example, ditches, ruts, potholes, accumulations of material 119 or other stockpiles 118, or other imperfections may exist along the working surface 110 of the worksite 112. Such imperfections may also exist within the loading path leading to one or more access points 210. Because navigating around these imperfections may cost the loader 104 time and fuel, any loading path that includes imperfections or obstacles may be eliminated to maintain the efficiency of the loader 104. By eliminating loading paths that include imperfections, the loader 104 may be more effective and efficient in moving material 119 within the stockpile 118. In one example, the controller 136 of the loader 104 may select a loading path that allows the loader 104 to navigate around or avoid imperfections while still being able to effectively scoop material 119 from the stockpile 118.

[0050] For example, in Figure 2 In the previous pass depicted in FIG, an entry point may have been selected that is between entry points 210-1 and 210-3 on the initial perimeter 202 of the stockpile 118, and as a result, a concave detent is formed in the stockpile 118 as shown by the dashed line 202-1. Figure 2 , a new entry point 210-2 may be identified when selecting a subsequent entry point 210 for the stockpile 118 depicted in FIG. 2 . However, in this subsequent pass, if the loader 104 selects the entry point 210-2, it may cause the stockpile 118 to be bisected, which may inevitably and undesirably result in the formation of two separate stockpiles of material 119. Therefore, when determining whether to select an entry point 210, the system controller 122 and / or the controller 136 of the loader 104 may exclude as a candidate any entry point 210 that would result in a bisection of the stockpile 118. Figure 2In the example, the system controller 122 and / or the controller 136 of the loader 104 may select entry points 210-1 and / or 210-3 because they are located on the side of the pile 118 (i.e., the front side 281) that is closest to the field of view of the loader 104 and / or the position sensor 130 and / or the perception sensor 134.

[0051] In the examples described herein, the number of candidate entry points 210 may be predetermined by utilizing the computing power of system controller 122 and / or controller 136 (including the ECM, which may be or may be included within controllers 122, 136). In one example, the predetermined number of candidate entry points 210 may be between 1 and 20. Furthermore, in one example, entry points 210 may be evenly distributed along the edge of pile 118 facing loader 104.

[0052] After determining the midpoint 204 and the plurality of cut-in points 510, a valid cut-in point 510 may be selected based on the criteria described above. Subsequently, for the selected cut-in point and / or for each candidate cut-in point, a plurality of candidate loading directions 212 may be identified by the system controller 122 and / or the controller 136. The candidate loading directions 212 may include any direction that begins at the cut-in point 210 and enters the stockpile 118, such that the direction cuts into the stockpile 118 rather than away from and / or out of the stockpile 118. Furthermore, the number of candidate loading directions 212 identified by the system controller 122 and / or the controller 136 may be predetermined by the computing power of the system controller 122 and / or the controller 136 (including the ECM, which may be or may be included within the controllers 122, 136). For example, the system controller 122 and / or the controller 136 may identify between one and ten candidate loading directions 212 for each cut-in point 210. Furthermore, similar to the entry point 210, the loading directions 212 may be evenly distributed throughout the angular range. For example, if three loading directions 212 are identified for the entry point 210, the angle between the first and second loading directions of the three loading directions 212 may be the same as the angle between the second and third loading directions of the three loading directions 212.

[0053] For a selected entry point 210 and loading direction 212, or for each identified entry point 210 and loading direction 212, the system controller 122 and / or the controller 136 of the loader 104 may evaluate a cost function associated with selecting a pair of entry point 210 and loading direction 212 compared to other pairs. In one example, the cost function may be defined as follows:

[0054] Cost = w1 - w2 + w3 + w4 Equation 2

[0055] Where w1 is the travel distance between the current position of the loader 104 and the entry point 210, w2 is the distance between the entry point 210 and the midpoint 204 of the stockpile 118 of material 119, w3 is the deviation between the loading direction 212 and the direction from the entry point 210 to the midpoint 204 of the stockpile 118 of material 119, and w4 is the integral of the steering force applied from the back-up point 255 to the entry point 210 of the stockpile 118.

[0056] In one example, the variables within Equation 2 are non-negative scalars. In this example, the non-negative scalars can be predetermined through historical analysis of existing customer data using, for example, machine learning techniques and / or linear regression. Machine learning uses algorithms and statistical models to enable the loader 104 to perform a specific task without continuous, explicit command input. The specific task learned here is to semi-autonomously or autonomously move material 119 within a pile 118 throughout the worksite 112 in a cost-effective manner. Cost-effectiveness may include consideration of the time spent moving the loader 104, the fuel consumed while moving the loader 104, wear and tear on the fuel engine and its components, and other cost-effective considerations, and these resources can be optimized and / or conserved. In this example, the machine learning can be performed by the system controller 122 and / or the controller 136 of the loader 104, and the data can be stored in an associated data storage device. The system controller 122 and / or the controller 136 of the loader 104 can rely on patterns and inferences regarding how to most efficiently move around the loader 104. The mathematical model may be constructed by the system controller 122 and / or the controller 136 of the loader 104 based on training data obtained from, for example, the operation of the loader 104 sensed by a human operator. This training data may be used as a basis for the system controller 122 and / or the controller 136 of the loader 104 to determine how to predict or decide to move the loader 104 without being explicitly programmed to perform the task of moving material 119 from the stockpile 118.

[0057] Linear regression involves mathematical techniques for modeling the relationship between a scalar response (i.e., a dependent variable) and one or more explanatory variables (i.e., independent variables). The relationship within linear regression is modeled using a linear prediction function whose unknown model parameters are estimated based on the data. The resulting linear model can be used to determine the conditional probability distribution of the response given the value of the predictor. In an example utilizing linear regression, historical actions, such as previously executed entry points 210 and loading directions 212 selected and executed by a human operator under similar circumstances, can be fitted to an approximate linear function. The regression can then be used to determine the weight of each variable within Equation 2. In one example, this linear regression technique can be performed offline, allowing other computing devices to perform this processing without utilizing computing resources associated with the system 100 or placing an undue burden on the system controller 122 and / or the controller 136 of the loader 104.

[0058] In the above example, the variable w4 may be set to 0 to ignore differences in wear on the steering system and / or tires of the loader 104. Based on the above example, the system 100 may select an entry point 210 and a loading direction 212 that correspond to the lowest cost function and use the selected entry point 210 and loading direction 212 to perform a pass within the material movement process performed by the loader 104. The above process may be performed each time the loader 104 performs a pass to obtain more material 119 from the stockpile 118.

[0059] Thus, the loading strategy defined by machine learning and / or linear regression methods not only considers travel distance but also penalizes actions that could disrupt the convex shape of the pile 118 of material 119 that is achieved and maintained during manual operation, minimizing the cleanup of unaggregated material 119. Thus, the loading procedures of a semi-autonomous or autonomously operated loader 104 can more closely mimic the operational expertise of a human operator. Furthermore, the strategies described herein can provide optimal entry point 210 and loading direction 212 selection for each pass, thereby minimizing bucket passes and minimizing the effort required to clean thin layers of residue or isolated small piles of residue.

[0060] In the description Figure 1 System 100 and Figure 2 Following the process described in Figure 3 and Figure 4 . Figure 3 It is depicted with Figure 1 and Figure 2, a flow chart of an example method 300 associated with the system 100 shown in FIG. The method 300 may include, at 301, sensing an exterior surface of a stockpile 118 of material 119, as described above, using sensors 130, 134. In one example, data obtained from the sensors 130, 134 may be used to detect points located along the surface of the stockpile 118. The sensors 130, 134 may transmit this data representing points along the surface of the stockpile 118 to, for example, a controller 136 of the loader 104 and / or the system controller 122.

[0061] At 302, method 300 may include determining, using a controller 136 and / or system controller 122 located on an at least partially autonomously controlled machine (e.g., loader 104), a midpoint 204 of a pile 118 of material 119 sensed by, for example, position sensors 130 and / or perception sensors 134. In one example, controller 136 may be assisted or controlled by system controller 122. In one example, position sensors 130 and / or perception sensors 134 may be used to sense the pile 118 and identify data representing coordinates or 3D positions along the surface of the pile 118. Sensors 130, 134 may transmit this data to controller 136 and / or system controller 122 of loader 104 for processing. In one example, the data representing coordinates or 3D positions along the surface of the pile 118 may be transmitted to controller 136 and / or system controller 122 of loader 104 via network 124, central station 108, and / or communication device 126.

[0062] At 304, controllers 122, 136 may determine multiple potential loading paths around midpoint 204 for loading machine 104. This determination may be based on data obtained from position sensors 130 and / or perception sensors 134. In the examples described herein, the loading paths may be determined based on path determination logic that considers multiple rules when identifying loading paths. In one example, a loading path may be considered if the end at the boundary of pile 118 includes a face of pile 118 that is convex relative to work tool 140 of loader 104. If the end at the boundary of pile 118 includes a face of pile 118 that is concave relative to loader 104, then loading material 119 at that point may cause pile 118 to bisect, or at least further bisect. Therefore, any loading paths that may result in a bisection of pile 118 may be eliminated as candidate loading paths. Conversely, loading paths that target a convex surface of pile 118 are less likely to result in such a bisection. Additionally, conservative loading paths that may result in reduced optimization and / or cost-effective movement of the loader 104 may also be eliminated as candidate loading paths. Other rules may be applied to determine loading paths, and the controllers 136, 122 may use these rules along with data obtained from the sensors 130, 134 to determine the boundaries of the stockpile 118 and identify loading paths that follow the specified rules along the boundaries.

[0063] The potential loading paths include at least one entry point 210, wherein at least one entry point 210 includes at least one loading direction 212. At 306, the controllers 122, 136 may select a primary loading path of the potential loading paths based on a cost function analysis. The controller 136 of the loader 104 and / or the system controller 122 performs the above-described cost function analysis based on at least one factor to analyze the effectiveness of the potential loading paths (including the entry point 210 and the loading direction 220) to obtain the most effective loading path. Similar to determining how to identify the loading paths, a number of rules may be applied when determining which loading path is most effective. For example, one rule may include determining which of the entry points 210 of the terminating loading paths is closer to the loader 104 when the loader is returning from the dump 250 and approaching the material pile 118 on the work surface 110 located at the work site 112 after leaving the back-up point 255. In another example, the rules may include determining which of a number of loading paths terminates at the entry point 210 located at the most prominent boundary of the stockpile 118 as determined by data from the sensors 130, 134 and the controllers 136, 122. Thus, the controllers 136, 122 may make this determination by applying defined rules.

[0064] At 308, the system controller 122 and / or the controller 136 of the loader 104 may cause the loader 104 to execute the loading instance defined by the master loading path. When the system controller 122 provides instructions to the loader 104, the system controller 122 may send a signal to the loader 104 via the network 124, the central station 108, and / or the communication device 126. The signal may include data defining actions for the loader 104 to perform in order to move the material 119 within the stockpile 118 using the most efficient loading path.

[0065] Figure 4 It is depicted with Figure 1 and Figure 2 . The method 400 may include, at 401, sensing an exterior surface of a stockpile 118 of material 119, as described above, using sensors 130, 134. In one example, data obtained from the sensors 130, 134 may be used to detect points located along the surface of the stockpile 118. The sensors 130, 134 may transmit this data representing points along the surface of the stockpile 118 to, for example, a controller 136 of the loader 104 and / or the system controller 122.

[0066] The method 400 may also include, at 402, determining the midpoint 204 of the pile 118 of material 119 sensed by a sensor (e.g., position sensor 130 and / or perception sensor 134) while the controller 136 is located on the at least partially autonomously controlled machine (e.g., loader 104 and / or system controller 122). In one example, the controller 136 may be assisted or controlled by the system controller 122. In one example, the position sensor 130 and / or perception sensor 134 may be used to image the pile 118 of material 119 and may transmit data representing the image of the pile 118 to the system controller 122 via the network 124, the central station 108, and / or the communication device 126.

[0067] At 404, controllers 122, 136 may determine a plurality of loading entry points 210 based on the ability of loader 104 to obtain a full load of material 119 within work tool 140 coupled to and actuated by loader 104. This determination may be based on data obtained from position sensors 130 and / or perception sensors 134, with controllers 122, 136 determining the amount of material 119 at that point in stockpile 118 that can fill the volume of work tool 140. Loading entry points 210 may include at least one entry point 210.

[0068] At 406, method 400 may include determining a plurality of loading directions 212 for each loading entry point 210 identified at 404. As described herein, a number of rules may be applied when determining which entry point is most effective. For example, one rule may include determining which entry point 210 terminating the loading path is closest to the loader 104 after the loader leaves the reverse point 255. In another example, a rule may include determining which of the plurality of entry points 210 is located at the most convex boundary of the pile 118, as determined by data from the sensors 130, 134 and the controllers 136, 122. Thus, the controllers 136, 122 may make this determination by applying defined rules.

[0069] Method 400 may also include determining at 408 whether controller 122, 136 has identified additional loading directions 220 for all loading entry points 210. In one example, controller 122, 136 may determine that one loading direction 220 (e.g., loading direction 212-2) may prevent the work tool 140 of loader 104 from capturing a full volume within the work tool 140 because it would pass over one side of the stockpile 118. In this example, selecting loading direction 212-10 or 212-11 may result in a full volume within the work tool 140. In response to determining that additional potential loading directions 220 for all loading entry points 210 (including all loading directions 220) are identified (408, a yes determination), method 400 may loop back to 406 and determine additional loading directions 220 for entry points 210. This looping back to 406 may occur any number of times to obtain a larger or more exhaustive number of loading directions 220 for each entry point 210.

[0070] In response to determining that the controllers 122, 136 have identified additional potential loading directions 220 for all loading entry points 210 (408, a yes determination), a second determination may be made at 410 as to whether additional loading entry points 210 have been identified. At 410, in response to determining that additional loading entry points 210 will be identified (410, a yes determination), the method 400 may loop back to 404 to determine the additional loading entry points 210. This looping back to 404 may occur any number of times before obtaining a larger or more exhaustive number of entry points 210. It is noted that, at 406, loading directions 220 for additional entry points 210 may also be identified. In this manner, all potential entry points 210 and their corresponding loading directions 220 are identified by the system controller 122 and / or the controller 136 of the loader 104. In response to determining that additional entry points 210 have been identified (410, a no determination), the method 400 may proceed to 412.

[0071] At 412, the system controller 122 and / or the controller 136 of the loader 104 may cause the loader 104 to perform a loading instance as defined by the selected primary entry point 210 and loading direction 220. Where the system controller 122 provides instructions to the loader 104, the system controller 122 may send signals to the loader 104 via the network 124, the central station 108, and / or the communication device 126. The signals from the system controller 122 and / or the controller 136 of the loader 104 may include data defining actions to be performed by the loader 104 in order to move the material 119 within the stockpile 118 using the primary loading path. Figure 4 The process 400 can be iteratively performed a number of times to move the material 119 within the stockpile 118 to the dump 250. Thus, an iterative determination of the midpoint 204 can be calculated in each iteration, e.g., primary, secondary, tertiary, etc. Furthermore, the loading path selected throughout the material movement operation can be based on the cost function analysis described herein and the rules applied to determine which entry point 210 and loading direction 220 are most cost-effective.

[0072] Industrial Applicability

[0073] The present disclosure describes systems and methods for semi-autonomous or autonomous movement of material via a loader 104, such as a wheeled or tracked loader, a front shovel, an excavator, a cable shovel, a stocker, or any other similar machine. The movement of material can be based on the identification of a center point 204, a plurality of entry points 210, and a loading direction 220 for each entry point 210. Such systems and methods can be used to more efficiently move material using a semi-autonomous or autonomous loader 104 without creating a situation where a pile 118 of material 119 is bisected or the pile 118 is dispersed during multiple passes to create smaller piles of material and a cleanup process is performed to capture the dispersed material.

[0074] Although various aspects of the present disclosure have been particularly shown and described with reference to the above examples, those skilled in the art will appreciate that various additional examples may be contemplated by modifying the disclosed machines, systems, and methods without departing from the spirit and scope of the disclosure. Such examples should be understood to fall within the scope of the present disclosure as determined by the claims and any equivalents thereof.

Claims

1. A method for autonomously moving material (119), comprising: sensing an outer surface (202) of a stockpile (118) of material (119) using a sensor (134); determining, using a controller (136) located on the at least partially autonomously controlled machine (104), a midpoint (204) of the stockpile (118) of material (119) based on the sensed exterior surface (202); Determining, using the controller (136), a plurality of potential loading paths about the midpoint (204) for loading the machine (104), wherein determining the plurality of potential loading paths includes determining a plurality of loading entry points, each loading entry point having a plurality of candidate loading directions, each potential loading path including a loading entry point and a candidate loading direction; selecting, using the controller (136), a primary loading path from among the plurality of potential loading paths based on a cost function analysis; as well as The controller (136) is used to cause the machine (104) to execute a load instance defined by the master load path.

2. The method according to claim 1, wherein: The pile (118) includes a proximal side (281) facing the machine (104) and a distal side (280) opposite to the proximal side (281). Potential loading paths exclude paths to the far side (280) of the stockpile (118); and Potential loading paths exclude paths that include obstacles.

3. The method according to claim 1, wherein: The cost function analysis is performed based on at least one factor, the at least one factor including distance traveled, time to travel the travel distance, steering force, angle of repose of the material (119), and distance traveled by a work tool (140), and Determining the plurality of potential loading paths includes determining the plurality of potential loading paths based on an ability to obtain a full load of the material (119) within a work tool (140) coupled to and actuated by the machine (104).

4. The method according to claim 1, further comprising: determining, with the controller (136), a plurality of loading entry points (210) based on the ability of the machine (104) to obtain a full load of the material (119) within a work tool (140) coupled to and actuated by the machine (104); Determining a plurality of loading directions (212) of the loading entry point (210) using the controller (136); determining, using the controller (136), whether an additional loading direction (212) for the loading entry point (210) is to be identified; In response to determining that an additional loading direction (212) for the loading entry point (210) is to be identified, identifying an additional loading direction (212) for the loading entry point (210); determining, using the controller (136), whether additional load entry points (210) are to be identified; and In response to determining that additional load entry points are to be identified (210), the additional load entry points are identified (210).

5. The method according to claim 1, wherein: The cost function analysis includes determining whether a potential loading path will result in a bisection of the stockpile (118) of material (119), and In response to determining that a first one of the plurality of potential loading paths will result in bisection of the stockpile (118) of material (119), the first one of the plurality of potential loading paths is removed from consideration as one of the plurality of potential loading paths.

6. A system (100) for autonomously moving material (119), comprising: An at least partially autonomous machine (104) configured to travel along a work surface (110) at a work site (112), the machine (104) including a work tool (140) configured to carry material (119) as the machine (104) travels along the work surface (110); a sensor (134) configured to sense an outer surface (202) of a stockpile (118) of material (119); and A controller (136) in communication with the sensor (134) and the machine (104), the controller (136) being configured to: determining a midpoint (204) of a stockpile (118) of material (119) based on the sensed exterior surface (202); Determining a plurality of potential loading paths around the midpoint (204) for loading the machine (104), wherein determining the plurality of potential loading paths includes determining a plurality of loading entry points, each loading entry point having a plurality of candidate loading directions, and each potential loading path including a loading entry point and a candidate loading direction; selecting a primary loading path among the plurality of potential loading paths based on a cost function analysis; and The machine (104) is caused to execute a load instance defined by the master load path.

7. The system of claim 6, wherein the sensor (134) is at least one of a position sensor (134), an imaging device, a light detection and ranging device, a radar device, a sonar device, and a satellite imaging device.

8. The system of claim 6, wherein the cost function analysis is performed based on at least one factor, the at least one factor comprising a distance traveled, a time to travel the distance traveled, a steering force, an angle of repose of the material (119), and a distance moved by the work tool (140).

9. The system of claim 6, wherein determining the plurality of potential loading paths comprises determining a plurality of loading directions (212) based on an ability to obtain a full load of the material (119) within a work tool (140) coupled to and actuated by the machine (104).

10. The system of claim 6, wherein determining the plurality of potential loading paths for loading the machine (104) comprises: determining, with the controller (136), a plurality of loading entry points (210) based on the ability of the machine (104) to obtain a full load of the material (119) within a work tool (140) coupled to and actuated by the machine (104); and A plurality of loading directions (212) of the loading entry point (210) are determined.

Images