Floor handling using autonomous mobile robots

Abstract

An embodiment relates to a method of controlling an autonomous mobile robot, comprising the following steps: controlling the robot in a processing mode to process a floor surface by means of a floor processing module of the robot; detecting a dirt sensor signal representing the contamination level of the floor surface by means of a dirt sensor provided on the robot, and changing the speed of the robot based on the dirt sensor signal.

Description

[0001] Divisional application This application is a divisional application of Chinese invention patent application No. 2018800738939, filed on November 16, 2018, entitled "Floor processing by means of an autonomous mobile robot". Technical Field

[0002] This specification relates to the field of autonomous mobile robots used for processing surfaces. Background Technology

[0003] In recent years, autonomous mobile robots have been increasingly used for tasks such as cleaning floors. The key is that a given surface is treated entirely by surface-handling devices (such as brushes) attached to the robot. Simpler devices operate without needing to create and use a map of the robot's work area; instead, the device moves randomly within the area to be cleaned. More complex robots use a map of the robot's work area, created by the robot itself and provided electronically. These systems allow them to remember surfaces that have already been treated.

[0004] Modern autonomous mobile robots, which use maps of their application area for navigation, attempt to employ processing patterns with the most systematic possible motion or handling styles during surface handling (e.g., cleaning). This motion or handling style must be adjusted according to the complex surrounding environment of the robot's application area (e.g., a furnished residence). Furthermore, the robot needs to be able to respond to unexpected situations, such as people moving within its application area, obstacles difficult to detect by the robot's sensors, or highly polluted areas.

[0005] The specific path planning (trajectory planning) depends on the processing style and the collision avoidance strategy used in the corresponding processing mode. The way unprocessed sections are handled also depends on the processing mode. That is, the characteristics of the processing mode used to process the floor surface are mainly: the motion style (e.g., zigzag, spiral, etc.) adopted by the robot when attempting to cover the floor surface, the movement speed, the collision avoidance strategy used, and the strategy for subsequent processing of unprocessed sections of the floor surface (e.g., those left behind due to obstacles). For example, various schemes (processing modes) for robot-assisted floor surface processing are described in patent DE 10 2015 119865 A1.

[0006] In summary, the purpose of this invention is to improve existing robot-implemented methods for autonomous mobile robots used to process surfaces (e.g., for cleaning floors), thereby increasing the efficiency of the robots. Summary of the Invention

[0007] The present invention provides a solution for achieving the above-mentioned objectives as a method according to any one of claims 1, 16, and 20, and an autonomous mobile robot according to claim 25. The embodiments and derivative solutions are the subject of the dependent claims.

[0008] An exemplary embodiment relates to a method for controlling an autonomous mobile robot, comprising the steps of: controlling the robot in a processing mode to process a floor surface via the robot's floor processing module; detecting a dirt sensor signal representing the level of contamination on the floor surface using a dirt sensor disposed on the robot; and changing the robot's speed based on the dirt sensor signal.

[0009] Another embodiment relates to a method comprising the steps of: controlling a robot in a processing mode to process a floor surface using the robot's processing module; and creating a processing map marking areas of the floor surface that have been processed, and determining, based on the processing map, which areas still need to be processed. The method further includes detecting dirt sensor signals representing the level of contamination on the floor surface using dirt sensors mounted on the robot, and marking areas in the processing map related to the current robot position as "processed" or unmarked based on the dirt sensor signals.

[0010] Another embodiment relates to a method comprising the steps of: controlling a robot in a processing mode to process a floor surface using the robot's processing module and creating an intensity map in which a measure of the processing intensity of the floor surface is specified for each location or area on the floor surface.

[0011] An autonomous mobile robot comprising a control unit is also proposed, wherein the control unit is adapted to cause the robot to perform one or more of the methods described herein. Attached Figure Description

[0012] Exemplary embodiments will now be described in detail with reference to the accompanying drawings. The drawings are not necessarily to scale, and the invention is not limited to the aspects shown. Rather, the drawings are intended to illustrate basic principles. Wherein: Figure 1 This illustrates an autonomous mobile robot within the scope of robotic applications.

[0013] Figure 2 An exemplary design of an autonomous mobile robot is illustrated with a block diagram.

[0014] Figure 3 The block diagram illustrates the functional coordination between the sensor unit and control software of an autonomous mobile robot.

[0015] Figure 4A flowchart illustrating an example of the method described in this article.

[0016] Figure 5 A-5E illustrates exemplary motion patterns of an autonomous mobile robot when processing a floor surface according to embodiments described herein. Detailed Implementation

[0017] Figure 1 An example of an autonomous mobile robot 100 for manipulating a floor surface is shown. The modern autonomous mobile robot 100 is map-based, meaning it accesses an electronic map of its application area. During its movement across this application area, the robot detects obstacles. Obstacles may be objects such as furniture, walls, doors, etc. However, the robot can also detect people or animals as obstacles. In the example shown, the robot 100 has identified portions of walls W1 and W2 of the room. Methods for creating and updating maps, and for determining the location of the autonomous mobile robot 100 within the robot's application area with respect to this map, are well known. For this purpose, SLAM (Simultaneous Localization and Mapping) methods can be used, for example.

[0018] Figure 2 The block diagram illustrates examples of various units (modules) of the autonomous mobile robot 100. A unit or module can be a standalone component or part of the software used to control the robot. A unit may have several sub-units. The software responsible for the behavior of the robot 100 can be executed by a control unit 150 of the robot 100. In the example shown, the control unit 150 includes a processor 155 adapted to execute software instructions contained in memory 156. Some functions of the control unit 150 can also be implemented, at least partially, using an external computer. This means that the computing power required by the control unit 150 can be at least partially transferred to an external computer, which may be accessible, for example, via a home network or the Internet (cloud).

[0019] The autonomous mobile robot 100 includes a drive unit 170, which may have, for example, an electric motor, a transmission, and wheels, enabling the robot 100 to (at least theoretically) approach every point within the application range. The drive unit 170 is adapted to convert commands or signals received by the control unit 150 into motion of the robot 100.

[0020] The autonomous mobile robot 100 also includes a communication unit 140 for establishing a communication connection 145 to a human-machine interface (HMI) 200 and / or other external devices 300. For example, the communication connection 145 may be a direct wireless connection (e.g., Bluetooth), a local wireless network connection (e.g., WLAN or ZigBee), or an internet connection (e.g., access to cloud services). For example, the HMI 200 can output information about the autonomous mobile robot 100 to the user in visual or auditory form (e.g., battery status, current work task, map information such as a cleaning map, etc.) and accept user commands for the work tasks of the autonomous mobile robot 100. Examples of the HMI 200 include tablet PCs, smartphones, smartwatches and other wearable devices, computers, smart TVs, or head-mounted displays. As a supplement or alternative, the HMI 200 can be directly integrated into the robot, allowing the robot 100 to be operated, for example, via buttons, gestures, and / or voice input and output.

[0021] Examples of external devices 300 include computers and servers that perform calculations and / or store data, external sensors that provide additional information, or other household appliances (such as other autonomous mobile robots) that can collaborate with and / or exchange information with the autonomous mobile robot 100.

[0022] The autonomous mobile robot 100 may have a work unit 160, particularly a processing unit for treating (e.g., cleaning) floor surfaces. Such a processing unit may include, for example, a suction unit that generates an airflow to collect dirt, a brush, or other cleaning device. Alternatively or as a supplement, the robot may be adapted to apply cleaning fluid to the floor surface and perform the treatment.

[0023] The autonomous mobile robot 100 includes a sensor unit 120 with various sensors, such as one or more sensors for acquiring information about the robot's surrounding environment within its application area, such as the location and extent of obstacles and other landmarks within the application area. Sensors used to collect information about the surrounding environment include sensors for measuring distances to objects (e.g., walls or other obstacles) in the robot's surrounding environment. Various sensors are known for this purpose, such as optical and / or acoustic sensors, which can measure distances using triangulation or the propagation time of emitted signals (triangulation sensors, 3D cameras, laser scanners, ultrasonic sensors, etc.). Alternatively or supplementarily, cameras can be used to collect information about the surrounding environment. This is particularly useful when observing an object from two or more locations, allowing for the determination of the object's (obstacle's) location and extent.

[0024] In addition, robots may have sensors for detecting (often accidental) contact (or collision) with obstacles. This can be achieved through accelerometers (which, for example, detect changes in the robot's velocity in the event of a collision), contact switches, capacitive sensors, or other tactile or touch-sensitive sensors. Furthermore, robots may have floor sensors (also called fall sensors) to identify edges in the floor, such as steps. Other common sensors in the field of autonomous mobile robots are those used to determine the robot's speed and / or distance traveled, such as odometry or inertial sensors (accelerometers, tachometers) for determining changes in the robot's position and motion, and wheel contact switches for detecting wheel contact with the floor.

[0025] In addition, the robot may have sensors for detecting the level of contamination on the floor surface. These sensors are referred to herein as dirt sensors. For example, such sensors can detect dirt received by the robot during cleaning. For instance, a suction robot has a channel through which air carrying dirt (e.g., dust) is drawn from the floor to a dirt receiving container. A dirt sensor, for example, can provide a measure representing the amount of dirt carried in the airflow flowing through the channel. Another type of dirt sensor, for example, can detect the vibration and agitation of heavier dirt particles (e.g., using a piezoelectric sensor). As an alternative or complementary solution, the amount of dirt in the airflow can be detected optically. For example, the contamination level can be determined directly in a camera image. Another approach is a combination of one or more light sources and one or more photosensitive receivers. The dirt particles contained in the airflow scatter the emitted light according to their number and size, thus varying the intensity of the light detected by the receiver. Yet another approach is to detect dirt directly on the floor surface. For example, the contamination level can be detected directly via a camera. As an alternative or supplementary solution, the floor surface can be illuminated using a light source, and the level of contamination can be detected based on the characteristics of the reflected light. For example, contamination caused by liquids can be identified based on the conductivity of the floor surface. These and other sensors used to detect the level of contamination on floor surfaces are well known and will not be discussed further here.

[0026] In a simple example, the measurement signal from the dirt sensor displays at least two states. Here, the first state indicates no contamination or a normal level of contamination, and the second state indicates a severe level of contamination (i.e., the measured contamination exceeds a threshold). These two states can be distinguished, for example, based on a threshold of detected dirt particles. In principle, more than two states can also be distinguished (e.g., "clean," "normal contamination," "severe contamination"), allowing for a finer classification of the robot's response to contamination.

[0027] The autonomous mobile robot 100 can be mounted on a base 110, where it can, for example, charge its energy storage device (battery). The robot 100 can return to the base 110 after completing its task. If the robot no longer needs to handle a task, it can wait for a new application in the base 110.

[0028] Control unit 150 is adapted to provide all the functions required by the robot to enable it to move independently within its application area and complete tasks. For this purpose, control unit 150 includes, for example, a processor 155 and a storage module 156 for executing software. Control unit 150 can generate control commands (e.g., control signals) for working unit 160 and drive unit 170 based on information obtained from sensor unit 120 and communication unit 140. Drive unit 170 is capable of translating these control signals or control commands into robot motion as described above. The software contained in memory 156 can also be modular. Navigation module 152 provides, for example, functions for automatically creating a map of the robot's application area and for path planning for robot 100. Control software module 151 provides, for example, general (global) control functions and can form an interface between the modules.

[0029] To enable the robot to autonomously complete tasks, the control unit 150 may include navigation capabilities within the robot's application area, provided by the navigation module 152 described above. These capabilities are well-known and may primarily include the following: • For example, but not only through SLAM methods, information about the surrounding environment is collected using sensor unit 120 to create (electronic) maps. • Manage one or more maps, with one or more robot application areas associated with these maps. • Based on environmental information measured by sensors in sensor unit 120, the robot's position and orientation (collectively referred to as "attitude") on the map are determined. • Map-based path planning (trajectory planning) from the robot's current pose (starting point) to the target point. • Contour following mode, in which the robot (100) moves along the contour of one or more obstacles (e.g., walls) at a substantially constant distance d from this contour. • Zoning identification, during which the map is analyzed and broken down into zoning zones, such as identifying spatial boundaries like walls and doors and passageways, so that these zoning zones describe the rooms of the residence and / or the reasonable partitioning of these spaces.

[0030] For example, when the robot's surrounding environment changes (obstacles move, doors open, etc.), the control unit 150 can continuously update the map of the robot's application area during robot operation with the help of the navigation module 152 and information based on the sensor unit 120.

[0031] In summary, the (electronic) map usable by robot 100 is a map dataset (e.g., a database) used to store location-related information about the robot's application area and the surrounding environment within that application area. In this context, "location-related" means that the stored information corresponds to various locations or poses on the map. That is, map data and map information are always based on specific locations or areas covered by the map within the robot's application area. In other words, a map represents a large dataset containing map data, and the map data can contain arbitrary location-related information. Here, location-related information can be stored at different levels of detail and abstraction, which can be matched to specific functions. In particular, information can be stored redundantly. A collection of several maps that cover the same area but are stored in different forms (data structures) is often referred to as a "map".

[0032] Based on (stored) map data, (currently measured) sensor data, and the robot's current task, navigation module 152 can plan the robot's path. At this point, it may be sufficient to determine waypoints (intermediate target points) and the target point. This plan can then be transmitted via control software module 151 in the form of specific drive commands. Based on these drive commands, drive unit 170 is controlled, and the robot thereby moves to the target, for example, along waypoints (from waypoint to waypoint). It should be noted that the planned path can include entire areas, processing trajectories, and / or short, direct movement segments (e.g., a few centimeters in the case of avoiding obstacles).

[0033] Autonomous mobile robots can have various operating modes for controlling the robot. Operating modes determine the robot's (internal and externally visible) behavior. For example, a robot that navigates using a map can have an operating mode for constructing new maps. Another operating mode can be provided for target point navigation; that is, the robot navigates from one point (e.g., the base) to a second point (the target point, such as the starting point of a task, particularly a cleaning task). Other operating modes (i.e., processing modes, particularly cleaning modes) can be provided to complete the robot's original task.

[0034] A robot may have one or more processing modes for handling floor surfaces. That is, to perform a specific (cleaning) task, the robot selects an operating mode based on specific criteria, and the robot operates in that operating mode during the task. In the simplest example, the user specifies the processing mode to be selected for the robot. Alternatively or additionally, a fixed sequence of processing modes may also be performed (e.g., surveying (sub-)areas, cleaning the edges of (sub-)areas, cleaning the surfaces of (sub-)areas). For example, in a strategy for covering a floor surface, the processing modes may vary. For example, the robot may be controlled randomly or systematically. Random control strategies typically do not use a map. Systematic migration strategies typically use a map (or a portion thereof) of the robot's application area, which may be constructed during processing or may be known before processing begins (e.g., from longer movement patterns or from previous applications). A typical example of a systematic migration strategy for covering a floor surface is based on motion patterns corresponding to each operating mode (also called processing / cleaning patterns during the processing / cleaning process). Commonly used motion patterns involve movement along parallel, interconnected trajectories (zigzag). Another motion style is movement along a spiral trajectory. Another motion style can be following the outline of a predefined area, which may be composed of, for example, real and virtual obstacles, to achieve the processing of approaching a wall.

[0035] As supplementary or alternative solutions, treatment modes can be differentiated by the selection and use of cleaning tools. For example, there may be an operating mode using a suction unit with high suction efficiency and a rapidly rotating brush (carpet cleaning mode). Alternatively, there may be another operating mode (hard floor cleaning mode) using reduced suction efficiency and a slower brush rotation. Furthermore, a wiping unit can be used on hard floors. Additionally, depending on the floor type (e.g., stone, wood), cleaning solutions (e.g., water) may be applied to the floor or not.

[0036] The operating mode (cleaning mode) can be modified based on the signal values ​​from sensors (dirt sensors) used to detect the level of contamination on the floor surface. For example, a dedicated operating mode (stain cleaning mode) can be activated for heavily contaminated locations to eliminate localized contamination. In this example, for instance, it could switch from an operating mode with a zigzag motion pattern to an operating mode with a spiral motion pattern (stain cleaning mode). However, switching between operating modes is disruptive to systemic cleaning because it increases the complexity of the methods used (especially for navigation and trajectory planning). Therefore, it is desirable to use a simpler method that is easily integrated into the systemic operating mode to take into account the identified level of severe contamination. The path planning method employed by the robot may depend on the current operating mode.

[0037] Therefore, the operating mode (cleaning mode) used for cleaning floors is characterized primarily by its motion pattern (e.g., zigzag, spiral, etc.), which, among other factors, the robot employs to attempt to cover the floor surface of current interest (e.g., a specific room or a portion thereof) as completely as possible, uses collision avoidance strategies (e.g., returning, avoiding obstacles, etc.), and employs strategies for subsequent processing of unprocessed areas of the floor surface (e.g., areas left behind due to obstacles). For example, after the motion pattern corresponding to the processing mode (e.g., a zigzag trajectory) has been executed, the remaining area can be approached and cleaned. Other approaches use a processing map or cleaning map, in which cleaned surfaces are marked for subsequent processing at a later time. The exemplary embodiments described herein are primarily intended to minimize mode switching during the cleaning process, especially when high levels of contamination are detected.

[0038] Responding to pollution by reducing speed - Figure 3 This illustrates how to integrate sensors (dirt sensors) for detecting the level of contamination on a floor surface into the architecture of an autonomous mobile robot. In the navigation module 152, information about the robot's surroundings (provided by the navigation sensor 121 included in the sensor unit 120) and updated map data (e.g., obstacle locations, etc.) using rangefinding methods, along with the robot's position (e.g., according to SLAM methods, see below), are used. Figure 4Step S1). For the range-based method, the sensor unit 120 of robot 121 may have an odometer 122 (e.g., wheel encoder, optical odometer, etc.). The robot's path planning is then updated according to the current operating mode (processing mode, cleaning mode, etc.). Path planning is based on the motion pattern corresponding to the current operating mode, the collision avoidance strategy used in the corresponding operating mode (e.g., return, movement along the contour of obstacles, etc.), and the strategy used for subsequent processing of remaining partitions. Path planning may include, for example, determining waypoints to the target, defining path segments, motion vectors, and / or other elements used to describe the path the robot traverses the robot's application area. For home robots (unlike large robots that move quickly, such as autonomous vehicles), the dynamic characteristics (especially speed and acceleration) of the robot as it traverses the path are typically ignored in path planning. Path planning updates may also include checking whether the robot is still on the pre-planned path. In the case of deviations (especially those greater than tolerances), it can be determined how the robot can return to the planned path. In addition, path planning updates may include checking whether the pre-planned path can be implemented without collisions. For example, this process can avoid obstacles that were not previously present or were not considered in the path planning. After the path planning is updated, the first step (updating map data and robot position) can be repeated in the navigation module 152.

[0039] The path planning results from navigation module 152 are forwarded to control software module 151, which creates drive commands for drive unit 170 according to preset rules. Drive unit 170 may consist, for example, of two independently driven wheels located on a single axle (differential drive). Such drives and their control for following a path are well known. For example, linear motion is generated when both wheels are driven at the same speed. Rotation about a center point between the two wheels is achieved when both wheels rotate in opposite directions at the same absolute speed. Other drive units 170, such as drives with wheels, chains, or legs, are well known.

[0040] During the generation of drive commands via the control software module 151, specific constraints should be observed. For example, the acceleration experienced by the robot cannot and / or is not permitted to exceed certain values. Another example is setting a maximum speed. These constraints may be specified by the components of the robot being used, but may also be specified by the robot's surrounding environment. For example, during the operation of the motors being used, maximum rotational speed and / or power should be observed to ensure sustained operation. For example, the maximum speed is reached during long straight-line travel. Lower speeds are typically reached during turning and / or obstacle avoidance. Movement speed can be reduced in particular depending on the required precision of the movement.

[0041] Furthermore, the control software module 151 may include safety-related functions. These safety-related functions can trigger the robot's response to installation-related events (detected hazards) (e.g., emergency braking, avoidance strategies, etc.). Possible safety-related events include, for example, a collision detected by a sensor (collision sensor 124, bumper), or a falling edge detected by another sensor (drop sensor 125, drop-sensor). Accordingly, a possible response to a detected collision or a detected falling edge is an immediate stop (emergency stop) of the robot. Furthermore, the robot can then move back a preset distance (approximately 1-5 cm) to establish a safe distance from the obstacle (or falling edge). The navigation module 152 does not need to concern itself with this standardized response to safety-related events. It is sufficient for the navigation module 152 to determine the robot's current position and the position of the obstacle detected by the sensor (bumper, or drop-sensor), and to use this information to adjust and / or re-determine the path.

[0042] To enable the robot to identify heavily contaminated areas when cleaning the floor and to treat these areas more intensively, the robot can be equipped with sensors (dirt sensor 126) to detect the level of contamination on the floor surface. A simple way to perform more intensive cleaning on the surface is to reduce the robot's speed, thus allowing for a longer treatment time on the contaminated surface. Here, the speed can be reduced directly during the generation of drive commands via the control software module 151. This eliminates the need to separately provide information about the contamination to the navigation module 152. No corresponding adjustments to the treatment mode, and in particular the treatment strategy, are required. This allows for a quick and direct response to heavily contaminated locations. In particular, this enables a relatively fast response, such as to safety-related incidents. Depending on a complementary or alternative approach, the robot's direction of travel, rather than its speed, can also be changed. For example, it can slightly retreat and then move forward again, thus repeatedly cleaning the floor area covered by this strategy. As a complementary or alternative approach, the currently planned trajectory can also be changed (and thus the direction of travel is also changed). For example, it could be modified as follows: the floor area identified as severely contaminated (through a "detour" caused by the changed trajectory) is repeatedly covered, and then the vehicle continues along the originally planned trajectory.

[0043] For example, in the case of detecting a severe pollution level, the speed can be adjusted so that the maximum allowed speed changes from a first value v1 to a second maximum speed v2. In particular, the maximum speed can be reduced (v2 < v1, for example v2 = 0.5·v1). In particular, this significantly reduces the speed during a long straight movement. For example, in an area where the robot needs to move more slowly due to an obstacle, the speed is only reduced if it is greater than the newly set maximum speed v2. This avoids unnecessary additional slowdown of the robot. For example, as long as the dirt sensor detects a severe pollution state, the reduced maximum speed v2 can be maintained. As an alternative or additional solution, after the severe pollution state switches back to the normal pollution state, the reduced maximum speed v2 can be maintained for a pre-set time (e.g., 5 seconds) or within a pre-set distance (e.g., 5 cm). Subsequently, the maximum speed is reset to its original value v1. This response to the increased pollution level does not affect the operation of the navigation module, especially not the path planning and / or the update of the path planning.

[0044] Another example of changing the speed in the case of detecting a severe pollution level is: when detecting a switch from the normal pollution state to the severe pollution state, the robot is stopped. For example, this can be done in a similar way to the response to a detected collision or a detected falling edge (e.g., sudden stop). This ensures a quick response to the severely polluted location.

[0045] As a supplement or alternative, the robot is capable of moving back. That is, the moving direction is reversed. For example, the robot can be directly controlled by a control software module to move back a pre-set distance (e.g., 5 cm). For example, this operation can be carried out in a similar way to the response to a detected collision or a detected falling edge. In particular, complex planning through the navigation module 151 is not required. As described above, based on the processing mode and basic processing strategies, and based on the robot's position and map information, the navigation module 151 can receive the path planning for the robot without taking into account the possible increased pollution level. The advantage of moving back is to compensate for the delay during the detection of severe pollution and to process the area with a potential increasing trend of pollution multiple times.

[0046] Before controlling the robot in the backward direction, the robot must stop. This can be achieved by a sudden braking strategy similar to a dangerous situation. As an alternative, this can be achieved by slow braking and (in the reverse direction) acceleration, thus creating a "softer" visual impression of forward and backward movement. A movement strategy similar to the response to a dangerous situation (stop and move backward) can also be used as a standard response. In this way, there is no need to adjust the processing mode implemented in the navigation module 152 for this either.

[0047] During the backward movement, the robot can move backward in a straight line or along the trajectory of its most recent movement. The latter can be achieved, for example, by inverting the most recently generated drive command. The distance or duration of the backward movement can be preset. Alternatively, sensor measurements can be used as conditions for stopping backward movement and resuming normal movement. For example, signals from a dirt sensor can be used. For instance, the robot can continue backward until the detection signal for the dirt level drops back below a preset threshold, or until an imminent collision with an obstacle. In one example, the robot continues backward until the dirt sensor no longer detects an increased level of dirt and continuously moves a defined distance (or duration). Collision avoidance can be activated during the backward movement.

[0048] In an alternative design, navigation module 151 can also receive information about the pollution level and perform path planning for the robot's backward movement. The advantage of this is that it can detect obstacles that may be behind the robot during robot control. Alternatively or supplementarily, backward movement can be controlled by control software module 152 as described above, wherein the movement is also monitored by a safety monitoring module. In cases of threat, such as collision with an obstacle or falling into a depression, the safety monitoring module can cause the movement to stop. This safety monitoring module can be a standalone module or part of control software module 151 and operate independently of navigation module 152.

[0049] After the robot stops and / or moves backward, it can resume forward movement. In this case, a reduced speed can be used, for example, for at least a preset distance or duration, as described above. Alternatively or supplementary, backward control of the robot can be activated each time severe contamination is newly detected, resulting in continuous reciprocating motion, similar to the method used by humans when dealing with severe contamination.

[0050] Responding to pollution levels by not marking them on the treatment map.- As a response to signals from dirt sensors, an alternative approach to controlling an autonomous mobile robot used for cleaning floor surfaces is to use a processing map (e.g., a cleaning map). In this map, all processed areas are marked. For example, this processing map can be displayed to a user, providing an overview of the robot's operations. Simultaneously, the robot can use this map to identify areas that still need processing. This allows, for example, the identification of areas that have not yet been processed due to the location of obstacles. When the robot passes by these unprocessed areas, (if the current processing pattern is interrupted) these areas can be included in the current processing. Alternatively, after processing areas according to the processing pattern (which depends on the operating mode), unprocessed areas can be identified based on the processing map, and the robot can be guided to process them. Such methods are well-known.

[0051] Figure 4 This illustrates an example of a scheme that controls an autonomous mobile robot based on signals from a dirt sensor without switching the processing mode on which the current processing strategy is based. Here, in the first step ( Figure 4 In step S1), in response to information provided by the navigation sensors regarding the robot's surrounding environment and the robot's range sensors, the map data concerning the surrounding environment and the robot's position are updated. In the second step ( Figure 4 In step S2), the processing map (e.g., a cleaning map) is updated. For this purpose, for example, the area between the robot's last known position and the position determined in the previous step S1 is marked as processed. The position of the processing unit (on the robot) may be taken into account. During the process of marking areas as processed (or unprocessed) in the processing map, data provided by the dirt sensor may be taken into account. The robot's path to be moved is then updated according to the current processing mode. For example, a strategy for further processing of leftover surfaces corresponding to the current processing mode could be: when the robot next passes by the leftover location, these surfaces are further processed, during which the robot moves according to the motion pattern used in the corresponding mode (e.g., a zigzag motion).

[0052] Here, when the dirt sensor detects no contamination or a normal level of contamination on a surface, the surface can be marked as processed. If a severe level of contamination is detected, the relevant surface is marked as severely contaminated in the processing map. Such marked surfaces will be processed multiple times. In the simplest case, the relevant surface is processed again by marking it as "unprocessed." Therefore, the marking of this surface is the same as the unprocessed area of ​​the floor surface. The purpose of this operation is that, depending on the processing strategy used for systematic or complete floor surface coverage, this area is identified as unprocessed, and the robot will be automatically guided back to this area in the future (based on the strategy used in the corresponding processing mode for subsequent processing of previously unprocessed surfaces). There is no need to modify or adjust the processing strategy and trajectory planning that directly considers detected severe contamination.

[0053] Figure 5 This shows an example of the processing order for the floor surface, and the corresponding markers in the processing map. Figure 5 A illustrates a robot 100 that systematically processes a floor surface using a meandering, interconnected trajectory. Areas marked "processed" on the processing map are shown with shaded lines. In the example shown, the robot is moving towards a locally heavily contaminated area D. Figure 5 In the scenario shown in B, robot 100 has reached the heavily contaminated area D, and thus detects this area using a dirt sensor.

[0054] In response to the detection of a severely contaminated area D, the robot's current location will not be marked as "processed." However, areas previously marked as "processed" can be remarked as "unprocessed" (or "pending further processing") for future processing. Figure 5 In the example shown in C, the area directly adjacent to or following robot 100 (e.g., with a fixed-defined width) is re-marked as "unprocessed". The advantage of this is that previously unidentified edge areas of heavily contaminated area D are also reprocessed. For example, a square with sides twice the robot's diameter and a center centered on the robot is marked as "unprocessed" (see [reference]). Figure 5 (A square drawn with a dashed line in C). It is important to note that areas marked as treated typically match the shape, size, and location of the processing unit in / on the robot. Similarly, areas marked as "untreated" due to identified severe contamination levels at least partially match the shape, size, and location of the processing unit.

[0055] Figure 5 D shows robot 100 on its next cleaning trajectory, arranged in a meandering pattern. The area D previously identified as heavily contaminated is marked "untreated" (see [reference]). Figure 5D (the square shown by the dotted line). Robot 100 identifies this based on the processed map. Accordingly, when it reaches this area D marked as "unprocessed", robot 100 processes the area again. Figure 5 E illustrates the possible processing patterns that arise from the reprocessing of region D. In alternative design schemes, the robot could also... Figure 5 As shown in Figure D, the vehicle travels along a straight path, then returns to the "untreated" area D at the end of a zigzag processing pattern, allowing for further processing of area D. These two processes significantly improve the cleaning effect in the heavily contaminated area D. The reason no special adjustments to the processing pattern are needed is due to the inherent characteristics of a map-based systematic processing strategy (identifying and processing residual areas).

[0056] Strength Map The aforementioned method for controlling autonomous mobile robots aims to perform more intensive cleaning on specific areas (especially those identified as heavily contaminated) than on other areas. If repeated high-intensity cleaning of specific areas is required, the valuable information obtained can be used for long-term optimization of the robot's application and to better adapt it to customer needs. Therefore, this information must be systematically recorded and analyzed.

[0057] Therefore, the first step is to record the actual local processing intensity in the map. This means recording whether all locations within the application area have been processed (if not, the processing intensity is zero), and at what intensity.

[0058] A measure of intensity can be, for example, the duration of treatment when the robot stops at a heavily contaminated location, or moves forward and then back. As a supplement or alternative, the frequency of treatment can be a measure of intensity, or influence intensity, when the robot passes through a heavily contaminated location multiple times. Similarly, speed can be a measure of intensity, or influence intensity, when the robot passes through a location at a reduced speed. Finally, the treatment efficiency during treatment can also be a measure of intensity, or influence intensity, when the robot's suction efficiency is increased, for example. If the surface is treated at a slower pace, the time the robot spends on that surface increases; if the surface is treated multiple times, the time the robot spends on the relevant surface also increases. Therefore, a measure of treatment (cleaning) intensity can be the product of the treatment time of a segment and the treatment efficiency used (e.g., the suction efficiency of the suction unit, in general: the possible amount of contaminant removed per unit time). This product (time multiplied by treatment efficiency) can also be considered as the "work" generated during the treatment of a unit area of ​​the floor surface.

[0059] A simple approach to creating such an intensity map is, for example, to store the robot's current position at regular intervals (e.g., once per second). This produces a map containing a point cloud. In areas where the robot appears more frequently and / or where the robot stays longer (e.g., due to reduced speed), the stored robot positions are more densely packed with points than in other areas. Therefore, the spatial density of robot positions stored in this way (points per unit area) is a usable measure of processing intensity.

[0060] The reason for the higher processing intensity (as mentioned above) may be: a response to the data provided by the dirt sensor, thereby allowing for multiple and / or slower processing of areas identified as heavily contaminated.

[0061] Another reason for the increased processing power may be the information provided by navigation sensors about the robot's surroundings. For example, speed may be reduced near obstacles, and especially in scenarios involving cleaning walls and / or corners. The advantage is that the reduced speed allows for more precise navigation. This enables cleaning closer to obstacles and around corners. Furthermore, cleaning efficiency is additionally improved, thus removing dirt accumulated in corners and edges more effectively.

[0062] Another reason for the increased processing power in the region is explicit user instructions. For example, users can (e.g., through a human-machine interface 200, see...) Figure 2 The user can instruct robot 100 to perform more intensive and / or repeated cleaning of the area it is in. Alternatively or as a supplement, the user can instruct the robot to perform a more thorough and intensive cleaning of a space (e.g., a passageway) or area (e.g., a dining area). For this purpose, the robot's map data can be displayed on an HMI (e.g., a tablet) as a floor plan of the robot's application area. The user can then directly mark areas requiring more intensive cleaning on the displayed map. For example, the user can select a cleaning program (processing mode) to perform a more intensive cleaning on the floor.

[0063] Furthermore, information stored in the map can prompt the robot to perform more intensive processing on a location or area. This could be information input by the user, such as explicit instructions to perform more intensive processing on a room (or a portion thereof). Alternatively or supplementary, the user can also use indirect information, such as room names (e.g., “kitchen”), area names (e.g., “entrance area”), and / or object names (e.g., “table”), to adjust the processing intensity. For example, the room name “kitchen” or the area name “entrance area” or “dining area” can indicate a particularly high cleaning need in that location.

[0064] Furthermore, the robot can learn information about the necessity of high-intensity treatments. For example, the robot can determine that an area with a higher level of contamination always requires a higher intensity of cleaning than other areas. To this end, for example, an intensity map can be stored after each treatment application. The stored map can be analyzed in terms of patterns and changes across several treatment applications. This allows, for example, the identification that a room (at least partially) requires a higher intensity treatment in almost every application. Based on this, the robot can independently, or with user confirmation, always treat the room, either entirely or partially, in a higher-intensity treatment mode. Alternatively, the user can be advised to clean the room more frequently. For example, if currently treated only every two days, daily treatment could be suggested.

[0065] As an alternative to storing the entire intensity map, it may be sufficient to determine and store areas with particularly high treatment intensities (e.g., due to severe pollution) after the treatment application. For example, storing areas and / or locations where the treatment intensity is greater than the minimum, corresponding to the average of the total intensity, and / or the intensity of the standard treatment mode.

[0066] This intensity map can be used as an alternative to processing maps and pollution maps (i.e., maps that correspond to the floor surface pollution levels measured by sensors for each location or area within the application area). As an alternative, this intensity map can serve as a useful supplement to the aforementioned maps, providing users with the necessary information more simply and directly, and improving the robot's autonomous learning capabilities.

Claims

1. A method for controlling an autonomous mobile robot (100), comprising the following steps: The robot (100) is controlled to process the floor surface in processing mode via its floor processing module (160). Using a dirt sensor (126) mounted on the robot, a dirt sensor signal representing the level of contamination on the floor surface is detected. During the treatment of the floor surface, the treatment mode is adjusted in response to the dirt sensor signal.

2. The method according to claim 1, wherein, The adjustment of the processing mode includes: During the processing of the floor surface, the robot's movement speed and / or movement direction are modified based on the dirt sensor signals.

3. The method according to claim 1, wherein, The adjustment of the processing mode includes: The motion pattern of the processing mode is changed in response to the dirt sensor signal.

4. The method according to claim 3, wherein, When a high level of pollution is detected, the motion pattern in the processing mode switches from a meandering motion pattern to a spiral motion pattern.

5. The method according to claim 1, wherein, The adjustment of the processing mode includes: In response to the dirt sensor signal, the anti-collision strategy is changed.

6. The method according to claim 1, wherein, The adjustment of the processing mode includes: In response to the dirt sensor signal, the strategy for processing previously untreated floor surface zones is changed.

7. The method according to claim 1, wherein, After completing the execution of the motion pattern associated with a specific processing mode, align and process the areas missed due to obstacles.

8. The method according to claim 1, wherein, A processing map is used to process the floor surfaces, in which the processed surfaces are marked.

9. The method of claim 8 further comprises reprocessing the surfaces marked as processed in the processed map.

10. A method for controlling an autonomous mobile robot (100), comprising the following steps: The robot (100) is controlled to process the floor surface in processing mode via its floor processing module (160). Using a dirt sensor (126) mounted on the robot, a dirt sensor signal representing the level of contamination on the floor surface is detected. The speed of the robot (100) is changed during the treatment of the floor surface based on the dirt sensor signal.

11. The method according to claim 10, The dirt sensor signal can present a first state and a second state according to the level of dirt on the floor surface.

12. The method according to claim 11, The first state of the dirt sensor signal indicates a normal level of contamination, and the second state of the dirt sensor signal indicates a severe level of contamination.

13. The method according to claim 10 or 11, The processing mode described therein corresponds to the robot's maximum speed, and The maximum speed mentioned above depends on the state of the dirt sensor signal.

14. The method according to any one of claims 11 to 13, In response to the second state of the dirt sensor signal, the robot's speed decreases from the first value to the second value.

15. The method according to claim 14, After the speed decreases, the speed is restored to the first value according to at least one preset standard.

16. The method according to claim 15, The at least one preset criterion includes at least one of the following: the dirt sensor signal re-presents the first state, the dirt sensor signal re-presents the first state and a preset time has elapsed since then; the dirt sensor signal re-presents the first state and the robot has moved a defined distance since then; a preset time has elapsed since the speed decreased, and the robot (100) has moved a preset distance since the speed decreased.

17. The method according to any one of claims 11 to 16, The robot (100) stops when the dirt sensor signal switches from the first state to the second state.

18. The method according to any one of claims 11 to 17, In response to the dirt sensor signal switching from a first state to a second state, the robot (100) moves back.

19. The method according to claim 18, The robot (100) moves backward a preset distance and / or duration along a straight line or the trajectory it followed when it came, thereby achieving the backtracking.

20. The method according to claims 18 and 19, The process of moving back takes obstacles into account to prevent collisions.

21. The method according to any one of claims 18 to 20, The robot (100) stores information about the location of obstacles in a map and uses the information stored in the map to avoid collisions during retraction, without using current sensor information about obstacles.

22. The method according to any one of claims 17 to 21, After stopping or reversing, the processing mode continues along the normal movement direction at a reduced speed.

23. The method according to any one of claims 10 to 14, in, In the processing mode, the robot (100) moves within a range on the floor surface at a speed less than or equal to the maximum speed corresponding to the processing mode, and wherein The maximum speed is reduced, thereby changing the speed of the robot (100).

24. The method according to any one of claims 10 to 23, wherein control of the robot in processing mode comprises: Based on the motion pattern corresponding to the processing mode, the obstacle avoidance strategy corresponding to the processing mode, and the strategy corresponding to the processing mode for subsequent processing of unprocessed surfaces, path planning is implemented based on map information and robot position. The planned path is converted into driving commands.

25. A method for controlling an autonomous mobile robot (100), comprising the following steps: The robot (100) is controlled to process the floor surface in processing mode via its processing module (160). A processing map is created, in which the processed areas of the floor are marked, and the processing map is used to determine areas that still need to be processed. Using a dirt sensor (126) mounted on the robot, a dirt sensor signal representing the level of contamination on the floor surface is detected. In response to the dirt sensor signal, during floor processing, an area (D) is marked or unmarked as "processed" based on the robot's current position in the processing map.

26. The method according to claim 25, The dirt sensor signal can present a first state and a second state according to the dirt level of the floor surface, and The first state of the dirt sensor signal indicates a normal level of contamination, and the second state of the dirt sensor signal indicates a severe level of contamination.

27. The method of claim 25 or 26, wherein, in response to the dirt sensor signal, marking the region based on the current robot pose in the processing map includes: If the dirt sensor signal indicates a severe level of contamination, prevent the area (D) associated with the current robot position from being marked as "processed," regardless of whether the area has been processed.

28. The method according to any one of claims 25 to 27, wherein the area (D) is marked as "treated" only if the area (D) associated with the current robot position has been actually treated and the dirt sensor signal indicates a normal level of contamination.

29. The method according to any one of claims 25 to 28, wherein if the area (D) has been actually treated and the contaminant sensor signal indicates a severe level of contamination, the area (D) is marked as "untreated" or "pending further treatment".

30. The method according to any one of claims 25 to 29, The region (D) that is relevant to the current robot position surrounds the robot's current position.

31. A method for controlling an autonomous mobile robot (100), comprising the following steps: In processing mode, the robot is controlled to process the floor surface. An intensity map is created during the treatment of the floor surface, in which different locations or areas on the floor surface are specified as measures of the treatment intensity of the floor surface.

32. The method according to claim 31, The intensity of floor treatment is measured by at least one of the following: treatment duration, treatment frequency, treatment speed, cleaning efficiency during treatment, and the product of treatment duration and treatment efficiency.

33. The method according to claim 31 or 32, During floor treatment, the treatment intensity is controlled based on one of the following features: a signal from a dirt sensor indicating the level of contamination on the floor, information related to the location of obstacles in the robot's surrounding environment, or user instructions.

34. The method according to any one of claims 31 to 33, During the processing of the floor surface, the robot (100) navigates according to a map of the robot's application area, wherein an intensity map and / or information based on the intensity map are stored together with the map of the robot's application area.

35. The method according to any one of claims 31 to 34, in, Based on the intensity map, suggestions are made to the user via the human-computer interface (200) to perform more frequent and / or higher intensity processing on the room and / or area.

36. An autonomous mobile robot (100) comprising a control unit (150) adapted to cause the robot to perform a method according to any one of the preceding claims.

Images