Map update for a household robot

Abstract

An autonomously moving robot (100) for a household (105), wherein the robot (100) comprises: - a sensor (120) for scanning the household (105) in the vicinity of the robot (100); - a processing device (115) for determining an occupancy grid (135) based on the scanning; - wherein the occupancy grid (135) comprises a plurality of cells (140); - wherein a cell (140) represents a section of the household (105); - wherein for the cell (140) the logit of an occupancy by an object (110) is specified; - wherein the logit is the natural logarithm of a chance that the cell (140) is occupied by an object (110); - wherein the chance comprises a probability of occupancy divided by a probability of the cell (140) not being occupied;- wherein the processing device (115) is configured to control the movement of the robot (100) in the household (105) such that a section whose associated cell (140) has a logit exceeding a predetermined threshold is not entered; - further comprising a communication device for receiving another occupancy grid (135) from another robot (100) in the household (105); wherein the processing device (115) is configured to integrate the further occupancy grid (135) into the existing one by adding the logit values ​​of corresponding cells (140) respectively.;

Description

The invention relates to a household robot. In particular, the invention relates to updating a navigation map of an autonomously movable robot for a household.A robot is configured to move and navigate independently in a household. In this case, the robot is intended to perform a predetermined utility function, for example cleaning a floor surface. For navigation, the robot is provided with an orientation map which it can generate itself, for example on the basis of scanning of a surrounding area. The orientation map can be embodied in particular as a topographical map or as an occupancy map or occupancy grid. The occupancy map provides information about which locations of the household can be reached for the robot and how these locations are relative to one another. Points that cannot be reached can be permanent, for example in the form of a wall, or local, for example in the form of a displaceable table or chair.For example, the document DE 10 2019 203 274 B3 describes an autonomously movable robot for a household, which robot determines an occupancy grid on the basis of a scanning of the household in the environment of the robot.A particular challenge in the household is to update the occupancy grid continuously in order to enable a high mobility of the robot. For this purpose, the occupancy grid should have a format that enables a quick and efficient incorporation of new information in order to map, in particular, temporary obstacles as currently as possible in the occupancy grid.An object underlying the present invention is to provide an improved technique for updating an occupancy grid of an autonomously movable robot for a household. The invention achieves this object by means of the subject matter of the independent claims. Dependent claims represent preferred embodiments.An autonomously movable robot for a household comprises a sensor for scanning the household in the vicinity of the robot; a processing device for determining an occupancy grid on the basis of the scanning; wherein the occupancy grid comprises a plurality of cells; wherein a cell represents a section of the household; wherein the logit of an occupancy by an object is indicated for the cell; wherein the logit relates to the natural logarithm of a chance that the cell is occupied by an object; wherein the chance comprises a probability of an occupancy divided by a probability of a non-occupancy of the cell; wherein the processing device is configured to control a movement of the robot in the household such that a section whose associated cell has a logit that exceeds a predetermined threshold value is not entered; wherein the robot comprises a communication device for receiving a further occupancy grid of another robot in the household; and wherein the processing device is configured to integrate the further occupancy grid into the present one by adding the logit values of mutually corresponding cells in each case.It has been recognized that an occupancy grid ("occupancy grid") with logit (or "log odds") values is particularly well suited to mapping the frequently changing circumstances of a household for an autonomous robot. The occupancy grid may be updated based on samples by different sensors. A section initially determined as occupied can be determined as free again by sufficiently many determination of a free state, so that an obstacle no longer present can be mapped in the occupancy grid. The logit has a range of definition between 0 and 1. the logit maps probabilities between 0 and 0.5 to a negative value, a probability of 0.5 to a value of zero, and a probability between 0.5 and 1 to a positive value.The occupancy grid can also be updated on the basis of information not determined locally by the robot but imported. An order in which occupancy information is maintained is not important in this case. Local findings can thus be combined in an improved manner with information collected and supplied from the outside.According to the invention, the robot comprises a communication device for receiving a further occupancy grid of another robot in the household; wherein the processing device is configured to integrate the further occupancy grid into the present one by adding the logit values of mutually corresponding cells in each case. The result of an addition can then be stored in the corresponding cell in the occupancy grid.The maintenance of received information can be performed quickly and simply by adding logit values of mutually corresponding cells. The processing device may perform the updating during an operation of the robot. Operation cannot be affected by an update being made. Navigation of the robot may be possible even if some cells of the local occupancy grid have already been updated and others have not yet. If a local and a received occupancy grid relate to mutually overlapping, but not identical, regions of the household, an existing occupancy grid can be extended with respect to its region on the basis of the received information or only received information relating to the locally known region of the household is used. This can be advantageous in the case of a short storage space for storing the occupancy grid.The other robot can be constructed identically or differently from the first robot. For example, an occupancy grid can be exchanged between a moving, a flying or a permanently installed other robot. A prerequisite for combining two occupancy grids is merely that cells assigned to one another can be found in the grids.The processing device can be configured to store the received occupancy grid if no mutually corresponding cells can be determined between the occupancy grids. This can be the case in particular if the occupancy grids relate to disjunctive regions of the household or the local occupancy grid is not yet sufficiently strengthened to reliably identify a position of the occupancy grids with respect to one another on the basis of structures of occupied or free areas. The stored occupancy grid can be maintained at a later time when further information is available in the local occupancy grid. This further information can be sensed by the sensor or received from another source.Finally, information from any of a plurality of sources can be combined with one another in order to provide an improved occupancy grid for controlling the robot. The different occupancy grids may have been created under different conditions. For example, a robot that must not enter a predetermined area of the household can nevertheless receive information about its occupancy from another robot that can enter or scan the area without entering it. The received information can be used advantageously immediately or later.The processing device can be configured to scan a section of the household by means of the sensor, to determine the logit of the occupancy of the section on the basis of the scan and to add the determined logit to a logit of a cell of the occupancy grid representing the section. Thus, the local occupancy grid can be successively improved on the basis of separate scannings of the robot.The robot can also comprise a communication device for providing the determined occupancy grid. The occupancy grid can be provided in particular to another robot in the same household. In connection with the above-mentioned embodiment, two robots can exchange their occupancy grids in both directions and thus cooperate cooperatively in an improved manner.The processing device may further be configured to determine a local occupancy grid on the basis of a plurality of samples and provide the local occupancy grid. It is thus possible to avoid information having to be provided or transmitted individually with respect to each individual scan. By accumulating samples in a local occupancy grid, an efficient reduction of information to be transmitted or processed by a receiver can be performed.The processing device can be configured to integrate the local occupancy grid into a global occupancy grid after the provision and to re-initialize the local occupancy grid. The movement and obstacle travel of the robot can be influenced by knowledge of the global occupancy grid. By using a local occupancy grid, an effort for updating it can be reduced. Initializing the local occupancy grid may be done with a predetermined probability value such as 0.5 to express an uncertainty in the occupancy.The global occupancy grid may be built like the local one by having cells each representing a portion of the household and carrying logit values relating to their occupancy. The logit values can be easily and efficiently improved by adding based on multiple samples, as more fully set forth herein. Samples may relate to a predetermined area of the home represented by multiple cells. A sample corresponding to multiple cells may also be called a temporary occupancy grid. The temporary occupancy grid can be merged with another occupancy grid by cell-by-cell addition. In one embodiment, the occupancy grid can be provided whenever a predetermined number of samples has taken that have flowed into the occupancy grid. The occupancy grid can also be provided periodically.The local occupancy grid can be provided to predetermined receivers and kept ready until it has been successfully transmitted to all predetermined receivers. A receiver, in particular a further robot in the household, therefore does not have to be constantly available for receiving new occupancy grids, but can be temporarily switched off or temporarily move out of a communication area, for example. A dedicated memory can be provided for an occupancy grid not yet delivered to all receivers.According to a further aspect of the present invention, a system comprises a first robot described herein, which is configured to receive an occupancy grid, and a second robot described herein, which is configured to provide an occupancy grid. Each robot can also support both possibilities. The system can comprise any desired plurality of robots which can cooperate cooperatively in the manner described in order to exchange information about the occupancy state of sections of the household with one another. Thus, all robots can benefit from the scannings of the respective other robots and move safely in the household. In particular, quickly or frequently changing occupancy states by mobile objects can thus be determined and distributed in an improved manner. Planning a route of a robot can be improved by being able to identify and use portions of the household that have become free or free in an improved manner.The system may also include a central location configured to receive an occupancy grid from one of the robots and provide it to another robot. The central station can buffer a received occupancy grid until all provided robots have received it. In addition, the central location can create a global occupancy grid, which can be provided, for example, to a robot newly added to the household. The central location is preferably arranged in the household and is configured to receive or provide information only locally. A central point outside the household can also be provided, for example in the form of a server or a service, for example in a cloud.According to yet another aspect of the present invention, a method for controlling a system having a plurality of robots comprises steps of controlling at least one of the robots in the household on the basis of a local occupancy grid; scanning the household by the at least one robot; determining a local occupancy grid on the basis of the scan; and providing the determined local occupancy grid to at least one of the other robots. The other robot can then be controlled on the basis of the provided as well as optionally also a local occupancy grid. The method can be carried out with any number of robots described herein, which are preferably all assigned to the same household.An occupancy grid can be provided by a robot directly to one or more robots or via a central location. The robots can also communicate with one another in the manner of a meshed network ("mesh network") in order to transmit an occupancy grid to a plurality of receivers.An occupancy grid can be digitally signed by the providing robot or the central location in order to identify its identity or origin. For this purpose, a checksum ("hash") of the occupancy grid can be created and appended to the occupancy grid by means of cryptographic methods. An occupancy grid that is not or incorrectly underscribed can be discarded by a robot. The communication of a robot can be effected in encrypted form. A robot can maintain a history of stored occupancy grids, for example, by storing their checksums. By comparing a checksum of a received occupancy grid with stored checksums, it is possible to prevent an occupancy grid from being inserted a number of times.The invention will now be described in more detail with reference to the accompanying figures, in which: FIG. 1 shows a robot in a household; FIG. 2 shows a system with multiple robots; and FIG. 3 shows a flow diagram of a methodis.FIG. 1 shows a robot 100 in a household 105. The robot 100 is autonomously movable in the household 105 and is preferably configured to carry out a predetermined function on the way or at a controlled location. The function is usually assigned to the household 105 and can comprise, for example, transporting or cleaning. The robot 100 shown is ground-supported and may have a drive device comprising wheels; other types of movement of the robot 100 are also conceivable.The household 105 is usually used by at least one person who is moving in the household 105 and can move an object 110 in the household. The object 110 may include a fixed boundary such as a wall or cabinet. However, movable objects 110 are also possible, for example a movable piece of furniture such as a table or a chair or a book, a cable or a carpet. A movable door can be regarded as a movable object in this sense just as a person or a pet. Generally, an object includes any type of object with which the robot 100 is not to collide.A section that the robot 100 is not intended to travel is commonly referred to as a prohibited or no-go area and may coincide with a occupied section or may be self-contained. A prohibited area may also be permanent or temporary in nature. Frequently, a prohibited area can still be scanned, at least up to a certain distance from the robot 100. In one embodiment, the no-go region in the occupancy grid is occupied with a probability of 1, the logit of which is infinite. Thus, the area is always considered to be occupied, so that the robot 100 does not enter it.The robot 100 preferably comprises a processing device 115, at least one scanning device 120, a storage device 125 and a communication device 130. A drive device that is controllable by the processing device 115 in order to move the robot 100 in the household 105 is not shown.The processing device 115 is usually formed by an embedded controller. The scanning device 120 is configured to scan a section of the household 105 in the environment of the robot 100 as contactless as possible and to determine whether the section is occupied by an object 110 or is free. For example, the sensing device 120 may include a camera, a LiDAR sensor, a radar sensor, or an ultrasonic sensor. The sensing device 120 may also include a collision or contact sensor for an object 110. The storage device 125 is configured to record an occupancy grid 135 that reflects an occupancy state of the household 105. The communication device 130 is configured for preferably wireless communication with another robot 100 or another location in the household 105.The occupancy grid 135 comprises a number of cells 140, which are usually organized in the manner of a matrix or a checkerboard, wherein one cell 140 represents a section of the household 105. In the illustrated embodiment, a floor surface of the household 105 is represented by a two-dimensional array of cells 140. A cell 140 carries a value representing a logit of a probability that an associated portion of the household 105 is occupied by an item 110. In the illustration of FIG. 1, a light cell is assigned a low logit value and a dark cell is assigned a high logit value. If a section occupancy state is not known, cell 140 may be initialized to 0.5.The values assigned to cells 140 each include logit values of an occupancy of an assigned section. The logit is a function known from statistics, which is defined as follows:The robot 100 can scan a section of the household 105 by means of the scanning device 120 and determine a probability p with which the section is occupied by an object 110. With respect to the probability p, the logit can then be determined. An important property of the logit is that a further logit value is produced by simply adding two logit values. Thus, a logit value stored in a cell 140 of the occupancy grid 135 may be updated by adding a certain logit value. Whether the section is deemed busy or idle may be determined by comparing a stored logit value to a predetermined threshold. Hysteresis may be applied in the determination.The robot 100 may be controlled not to enter a portion represented by a cell 140 whose logit value is above a predetermined threshold. During driving, the robot 100 may update the occupancy grid 135 using samples.It is proposed that the robot 100 is configured to exchange an occupancy grid 135 or a section thereof with another robot 100 in the household 105. For this purpose, the robot 100 can provide the occupancy grid 135 determined by it to another robot 100 or unite an externally provided occupancy grid 135 with its local one.It is particularly preferred that the robot 100 holds in its storage device 125 a global occupancy grid 135, after which it determines its movement in the household 105, and a local occupancy grid 135 comprising information formed on the basis of one or more scans by means of the scanning device 120. At predetermined intervals, for example, each time a predetermined number of scans have taken place or after a predetermined time has elapsed, the local occupancy grid 135 can be provided to the outside and integrated into the own global occupancy grid 135. In another embodiment, each determination may also be incorporated directly into the global occupancy grid 135 while being simultaneously collected and periodically provided in the local occupancy grid 135. In yet another embodiment, the local occupancy grid 135 may also be provided after each sampling has taken place. This allows very little latency until the information is available at another robot 100, but also requires a high bandwidth for transmission.FIG. 2 shows a system 200 having a plurality of robots 100 in a household 105. By way of example, a first robot 205, a second robot 210 and a third robot 215 are illustrated. In addition, an optional central point 220 is provided by way of example, which is configured to communicate with the robots 205- 215 by means of a communication device 225, to process received information by means of a processing device 230 and to store it by means of a storage device 235. The robots 205- 215 can preferably share information about occupancy states of sections of the household 105 among each other. Various scenarios are conceivable for this purpose.In a first exemplary scenario, the first robot 205 is traveling in the home 105, while the second robot 210 is standing in a first charging station 240 and the third robot 215 is standing in a second charging station 245. Robots 210 and 215 may be powered down or operate at reduced power so that they are less readily communicatively accessible. In this scenario, the first robot 205 may build a local occupancy grid 135 without simultaneously providing it. Only when the trip is ended can the local occupancy grid 135 be provided to the robots 210, 215, so that these can work in all the current information in their occupancy grids 135. It is thus possible to prevent the robots 210, 215 from having to become active themselves more frequently or longer than necessary in order to update their occupancy grids 135.In a second exemplary scenario, robots 205 and 210 are traveling in the home while third robot 215 is stationary. By way of example, it is assumed that initially all robots 205- 215 have newly initialized occupancy grids 135, and that none of robots 205- 215 know its position or that of another robot 205- 215 on occupancy grid 135. The moving robots 205 and 210 can set up their occupancy grids 135 by communicating their own findings to one another immediately or after setting up a local occupancy grid 135. If the robots 205 and 210 end their journeys in the household 105, the last moving robot 205, 210 can provide its occupancy grid 135 to all other robots 205- 215, since it has the most up-to-date occupancy data.In a third exemplary scenario, all robots 205- 215 have an occupancy grid 135, and each robot 205- 215 knows its own position in the navigation map and the associated own local occupancy grid 135. If only one robot 205 now travels through the household 105, as in the first scenario, the latter can create a local occupancy grid 135. At the end of its travel, the local occupancy map 135 may be integrated into the global occupancy grid 135. Periodic transmission of the local occupancy grid 135 may not be necessary if the receiving robots 210, 215 are not currently using the occupancy grid 135.Since it is possible for the first robot 205 to operate in a different environment during the restarted journey than during a previous use, so that a local occupancy grid 135 does not correspond to the previously present global occupancy grid 135, it must be checked before the first addition of the temporary occupancy grid 135 whether the first robot 205 can locate itself in the present global occupancy grid 135 or whether the update of the temporary occupancy grid 135 to be newly integrated is to be assigned to the previous global occupancy grid 135. If this is the case, the first robot 205 supplies updates to the present global occupancy grid 135, otherwise, as in scenario 1, it creates a new global occupancy grid 135.As in the first scenario, the non-active robots 210, 215 are not charged with the receiving and processing of data in their idle phase and do not receive the global occupancy grid 135 as long as the first robot 205 is still moving. Since these robots 210, 215 are at their charging station 240, 245 all the time they do not require the updates of the occupancy grids 135 either. Instead, the first robot 205 sends the updated global occupancy grid 135 to the other robots 210, 215 after ending the trip, so that they can take over the occupancy grid 135 once. As in the second scenario, the robots 210, 215 must now respectively check whether the received occupancy grid 135 is an update of their existing global occupancy grid 135, or an alternative, new global occupancy grid 135, which is not intended to overwrite the existing global occupancy grid 135. In this way, all collected occupancy grids 135 in the most up-to-date version are available at the end of all robots 205- 215.In a fourth scenario, all robots 205- 215 have an occupancy grid 135, and each robot 205- 215 knows its own position with respect to the navigation map and the associated occupancy grid 135. Similar to the second scenario, robots 205 and 210 then move while third robot 215 is stationary. The moving robots 205, 210 respectively expand or update their occupancy grids 135 by creating a local occupancy grid 135 on the basis of scans of their environment and periodically maintaining it in the global occupancy grid 135. Since the local occupancy grid 135 is useful for the other moving robot 210, it is preferably also provided to it. The robot 210 can integrate the received occupancy grid 135 into its own global occupancy grid 135 and optionally provide its local occupancy grid 135 to the first robot 205 in a corresponding manner. Last time, each robot 205, 210 sends its temporary occupancy grid 135 to the other robot 205, 210 when it ends its travel.The non-active robot 215 receives an update from the last still moving robot 205, 210 that includes all the information that the moving robot 205, 210 has collected within its last mission.A prerequisite for overwriting the global occupancy grid 135 of the other robots 210, 215 is that they establish a correspondence with their own global occupancy grid 135. If no matches are found, robots 205-215 may act in a dressing but have been used in different environments. In this case, the global occupancy grid 135 of the last active robot 205, 210 is preferably stored as an alternative occupancy grid 135 on the other robots 205, 210, 215. The robot 205, 210 respectively driven in front of it sends its created global occupancy grid 135 to all other robots 210, 215 in this case until each of the collected global occupancy grids 135 could finally be assigned to one another or is stored as a new alternative occupancy grid 135 on all robots 205- 215. In this way, all of the robots 205-215 at the end have available to them all of the collected occupancy grids 135 in the most up-to-date version in each case.As a result of the described cooperation, all robots 205- 215 can have the most up-to-date occupancy grids 135 with the latest updates in order to base them on their own navigation. The described procedure is also applicable if the robots 205-215 of the dressing perform different tasks in the same household 105, for example suction, wiping, transport or a care task. Shorter exploration times and thus a faster and more efficient task performance may result if multiple robots 205- 215 explore cooperatively in the same household 105. Through the cooperative exchange of updates or extensions of the occupancy grids 135, those robots 205- 215 that are not driving also obtain the latest occupancy grids 135.Advantageously, an exchange of occupancy grids 135 can also take place from other environments, for example when the robots 205- 215 are en route in different rooms or on different floors. Even if a robot 205- 215 has not yet traveled there beforehand, it can then be used immediately there without exploration, since the most up-to-date occupancy grids 135 are already present to him. A user needs to create customized areas such as a prohibited area or additional information such as designation of rooms, do-not-jam times, or situations for specific rooms only once for the global occupancy grid 135, and not for each robot 205- 215 individually.FIG. 3 shows a flow diagram of a method 300 for transmitting map information from a first exemplary robot 205 to a second exemplary robot 210 that are in the same household 105. Steps associated with the first robot 205 are shown in a left-hand region and steps associated with the second robot 210 are shown in a right-hand region. It should be understood that the robots 205, 210 may also swap their roles. A bidirectional exchange is also possible. In addition, map information may be exchanged between any robots 205- 215 in a household 105, and any number of robots 205- 215 may participate in the method 300 shown. The example scenarios outlined above with reference to FIG. 2 are compatible with the method 300.In a step 305, the first robot 205 scans its environment in the household 105. A scan may relate to one or more sections to which different cells 140 of the occupancy grid 135 are associated. In a step 310, the occupancy grid 135 may be updated or expanded based on the sampling. The updating usually takes place on a local occupancy grid 135, while a navigation of the first robot 205 is influenced by a global occupancy grid 135.In a step 315, it is determined whether sufficient detections have been made. If this is not the case, steps 305 to 315 can be repeated. Otherwise, the created local occupancy grid 135 can be provided to one or more other robots 210, 215 in a step 320. If it is known which robots 210, 215 are to receive the update, a list can be created with identifications of all those robots 210, 215 that have not yet confirmed the receipt of the update. If a confirmation of a robot 210, 215 occurs, the identification can be removed from the list. In a step 325, it is determined whether the list is empty. If this is the case, the occupancy grid 135 can be discarded or even archived in a step 330. In addition, a new local occupancy grid 135 can be created, which is provided with new information during a new pass of steps 305, 310. The new occupancy grid 135 can also already be created and filled if not all receiving robots 210, 215 have acknowledged the reception yet.The second robot 210 can receive the occupancy grid 135 from the first robot 205 in a step 340. In a step 345, it can confirm the reception at the first robot 205. In a step 350, he can check whether the occupancy grid 135 is already known to him. For this purpose, it can determine a checksum over the received occupancy grid 135 and compare it with a list of checksums that were determined on the basis of previously obtained occupancy grids 135. In addition, the second robot 210 may check whether it may correlate the received information with the own information. If it cannot determine a position of the received occupancy grid 135 with respect to its own global occupancy grid 135, it can discard the received grid or retain it as an alternative occupancy grid 135. A subsequently received occupancy grid 135 can then be checked with respect to all stored alternative occupancy grids 135.Reference numerals denote reference numerals100 Robot 105 Household 110 Object, object, person, obstacle 115 Processing device 120 Scanning device, sensor 125 Storage device 130 Communication device 135 Occupancy grid, occupancy map, occupancy grid 200 System 205 First robot 210 Second robot 215 Third robot 220 Central location 225 Communication device 230 Processing device 235 Storage device 240 First charging station 245 Second charging station 300 Method 305 Environment scan 310 Local occupancy grid improve 315 Sufficient detections? 320 Local occupancy grid provide 325 All receivers confirmed? 330 Discard or archive 335 New local occupancy grid create 340 Local occupancy grid receive 345 Occupancy grid confirm 350 Occupancy grid known? 355 Received and local occupancy grid integrate

Claims

Autonomously movable robot (100) for a household (105), wherein the robot (100) comprises: - a sensor (120) for scanning the household (105) in the vicinity of the robot (100); - a processing device (115) for determining an occupancy grid (135) on the basis of the scanning; - wherein the occupancy grid (135) comprises a plurality of cells (140), - wherein a cell (140) represents a section of the household (105), - wherein for the cell (140) the logit of an occupancy is indicated by an object (110); - wherein the logit relates to the natural logarithm of a chance that the cell (140) is occupied by an object (110); wherein the chance comprises a probability of an occupancy divided by a probability of a non-occupancy of the cell (140); - wherein the processing device (115) is configured to control a movement of the robot (100) in the household (105) such that a section whose associated cell (140) has a logit that exceeds a predetermined threshold value is not entered, - further comprising a communication device for receiving a further occupancy grid (135) of another robot (100) in the household (105); wherein the processing device (115) is configured to integrate the further occupancy grid (135) into the present one by adding the logit values of mutually corresponding cells (140) in each case.Robot (100) according to claim 1, wherein the processing device (115) is configured to store the received occupancy grid (135) if no mutually corresponding cells (140) between the occupancy grids (135) can be determined.Robot (100) according to one of the preceding claims, wherein the processing device (115) is configured to scan a section of the household (105) by means of the sensor, to determine the logit of the occupancy of the section on the basis of the scan and to add the determined logit to a logit of a cell (140) of the occupancy grid (135) representing the section.Robot (100) according to claim 3, further comprising a communication device (130) for providing the determined occupancy grid (135).Robot (100) according to claim 4, wherein the processing device (115) is configured to determine a local occupancy grid (135) on the basis of a plurality of scans and to provide the local occupancy grid (135).Robot (100) according to claim 5, wherein the processing device (115) is configured to integrate the local occupancy grid (135) into a global occupancy grid (135) after the provision and re-initialize the local occupancy grid (135).The robot (100) according to claim 5 or 6, wherein the local occupancy grid (135) is provided to predetermined recipients and kept ready until it has been successfully transmitted to all predetermined recipients.A system (200) comprising a first robot (100) according to any one of the preceding claims and a second robot (100) according to any one of claims 3 to 7.The system (200) of claim 8, further comprising a central location configured to receive an occupancy grid (135) from one of the robots (100) and provide it to another robot (100).Method (300) for controlling a system (200) having a plurality of robots (100) in each case according to one of Claims 1 to 7, wherein the method comprises the following steps: - controlling at least one of the robots (100) in the household (105) on the basis of a local occupancy grid (135); - scanning the household (105) by the at least one robot (100) and determining a local occupancy grid (135) on the basis of the scanning; and - providing the determined local occupancy grid (135) to at least one of the other robots (100).

Images