DE102020216317A1 - Indoor localization of moving objects - Google Patents

Abstract

The present invention relates to a device (16) for determining a position of a moving object (12) within a previously known environment (14), having: a distance interface (24) for receiving distance data from a distance sensor (18) with information on distances of the moving object Object (12) from surrounding objects in the environment (14) within a field of view of the distance sensor; a first localization unit (26) for determining a first position estimate of the moving object (12) based on the distance data, a previously known map of the surroundings (14) and a previous position estimate of the moving object (12); a signal strength interface (28) for receiving signal strength data with information on signal strengths of radio signals from a plurality of transmitters (22a-22g) in the vicinity (14); a communication interface (30) for receiving reference data from a reference database (20) with information on signal strengths of radio signals from the plurality of transmitters (22a-22g) at previously known positions in the area (14); a second locating unit (32) for determining a second position estimate of the moving object (12) based on a comparison of the signal strength data with the reference data when an accuracy of the first position estimate is below a threshold value; and an evaluation unit (34) for outputting a final position estimate, the evaluation unit being designed to determine the final position estimate based on the second position estimate, the distance data, the previously known map of the surroundings (14) and the previous position of the moving object (12), if an accuracy of the first position estimate is below the threshold value, and configured to output the first position estimate as a final position estimate if the accuracy of the first position estimate is above the threshold value. The present invention further relates to a system (10) and a method.

Description

The present invention relates to a device for determining a position of a moving object within a previously known environment. The present invention further relates to a system for locating moving objects and a method for determining a position of a moving object and a computer program product.

The localization of moving objects, such as people or vehicles, is of great relevance in various fields of application. The term localization is understood to mean finding or determining a position of the object in relation to a more or less known environment. For example, it is absolutely necessary for autonomous vehicles in industrial environments to be able to localize them at any time in order to head for destinations and fulfill tasks based on this.

There are a number of different approaches to indoor localization. On the one hand there are feature-based localization methods. This includes, for example, 2D/3D scan matching based on lidar or camera data. A contour of the current environment is recognized, which is then used to locate the agent using a previously known map or without map information. However, this approach has the disadvantage that it is often not possible to achieve globally unambiguous results, since there can be several similar contours within the same environment (e.g. identical machining centers in a production hall). On the other hand, there are approaches in which tags or features are deployed within the environment, which are recognized by the agent (e.g. based on RFID or camera) and whose position is known. However, this approach has the disadvantage that these features or tags must be deployed first, requiring additional infrastructure customization. In addition, changes in the environment can occur, especially in dynamic environments, in which case the position of the tags must be remeasured or updated. Finally, there are signal-based localization methods in which the position of the agent is determined using received signals (WiFi, Bluetooth, Ultra-Wide-Band, etc.) via the signal strength and/or the signal propagation time. A disadvantage of these approaches is that either complex infrastructure adaptations are necessary, for example by deploying Bluetooth tags, or that only a low level of accuracy can usually be achieved when using the existing infrastructure.

In this context, the WO 2019/136019 A1 the mapping and control of a mobile cleaning robot. A mobile cleaning robot includes a drive system configured to navigate within an operating environment, a distance measuring device configured to communicate with other distance measuring devices within the operating environment, and processors that receive distance measurements from the distance measuring devices in the environment. Each distance measurement represents a distance between the mobile cleaning robot and corresponding devices in the area.

In this context, the object of the present invention is to provide a reliable approach for locating moving objects. In particular, a localization of moving objects in a previously known interior environment, in which the floor plan may change during operation, should be made possible. Precise localization should be achieved with as little effort as possible for adapting the existing infrastructure.

• einer Abstandsschnittstelle zum Empfangen von Abstandsdaten von einem Abstandssensor mit Informationen zu Abständen des beweglichen Objekts von Umgebungsobjekten in der Umgebung innerhalb eines Sichtfelds des Abstandssensors;
• einer ersten Lokalisierungseinheit zum Ermitteln einer ersten Positionsschätzung des beweglichen Objekts basierend auf den Abstandsdaten, einer vorbekannten Karte der Umgebung sowie einer vorherigen Positionsschätzung des beweglichen Objekts;
• einer Signalstärkenschnittstelle zum Empfangen von Signalstärkedaten mit Informationen zu Signalstärken von Funksignalen mehrerer Sender in der Umgebung;
• einer Kommunikationsschnittstelle zum Empfangen von Referenzdaten von einer Referenzdatenbank mit Informationen zu Signalstärken von Funksignalen der mehreren Sender an vorbekannten Positionen in der Umgebung;
• einer zweiten Lokalisierungseinheit zum Ermitteln einer zweiten Positionsschätzung des beweglichen Objekts basierend auf einem Vergleich der Signalstärkedaten mit den Referenzdaten, wenn eine Genauigkeit der ersten Positionsschätzung unterhalb eines Schwellenwerts liegt; und
• einer Auswerteeinheit zum Ausgeben einer finalen Positionsschätzung, wobei die Auswerteeinheit zum Ermitteln der finalen Positionsschätzung basierend auf der zweiten Positionsschätzung, den Abstandsdaten, der vorbekannten Karte der Umgebung sowie der vorherigen Position des beweglichen Objekts ausgebildet ist, wenn eine Genauigkeit der ersten Positionsschätzung unterhalb des Schwellenwerts liegt, und zum Ausgeben der ersten Positionsschätzung als finale Positionsschätzung ausgebildet ist, wenn die Genauigkeit der ersten Positionsschätzung oberhalb des Schwellenwerts liegt.

To solve this problem, the present invention relates in a first aspect to a device for determining a position of a moving object within a previously known environment, with:

• a distance interface for receiving distance data from a distance sensor with information on distances of the moving object from surrounding objects in the environment within a field of view of the distance sensor;
• a first localization unit for determining a first position estimate of the moving object based on the distance data, a previously known map of the surroundings and a previous position estimate of the moving object;
• a signal strength interface for receiving signal strength data including information on signal strengths of radio signals from a plurality of transmitters in the vicinity;
• a communication interface for receiving reference data from a reference database with information on signal strengths of radio signals from the plurality of transmitters at previously known positions in the area;
• a second localization unit for determining a second position estimate of the moving object based on a comparison of the signal strength data with the reference data when an accuracy of the first posi tion estimate is below a threshold; and
• an evaluation unit for outputting a final position estimate, the evaluation unit being designed to determine the final position estimate based on the second position estimate, the distance data, the previously known map of the area and the previous position of the moving object if an accuracy of the first position estimate is below the threshold value , and is designed to output the first position estimate as a final position estimate if the accuracy of the first position estimate is above the threshold value.

• einer Vorrichtung wie zuvor definiert;
• einem Abstandssensor an dem beweglichen Objekt zum Empfangen von Abständen von Umgebungsobjekten;
• einer Referenzdatenbank zum Speichern von Referenzdaten und zum Aktualisieren der Referenzdaten basierend auf übermittelten Signalstärkedaten und finalen Positionsschätzungen; und
• mehreren Sendern zum Aussenden von Funksignalen.

In a further aspect, the present invention relates to a system for locating moving objects within a previously known environment, comprising:

• a device as previously defined;
• a distance sensor on the moving object for receiving distances from surrounding objects;
• a reference database for storing reference data and for updating the reference data based on transmitted signal strength data and final position estimates; and
• several transmitters for sending out radio signals.

Further aspects of the invention relate to a method designed in accordance with the device and a computer program product with program code for carrying out the steps of the method when the program code is executed on a computer. In addition, one aspect of the invention relates to a storage medium on which a computer program is stored which, when executed on a computer, causes the method described herein to be carried out.

Preferred developments of the invention are described in the dependent claims. It goes without saying that the features mentioned above and those still to be explained below can be used not only in the combination specified in each case, but also in other combinations or on their own, without departing from the scope of the present invention. In particular, the device, the system, the method and the computer program product can be implemented in accordance with the configurations defined for the device and the system in the dependent claims.

According to the invention, distance data, ie sensor data from a distance sensor, is first received and the position of the moving object is estimated on this basis. The distance sensor is arranged on the moving object to be localized. Starting from a previously known map of the surroundings, localization can be carried out by evaluating the measured distances to the surrounding objects, ie all objects within the field of view of the distance sensor. If this localization does not have sufficient accuracy, for example due to ambiguity, localization is carried out based on signal strength data. For this purpose, signal strengths of radio signals from a number of transmitters in the vicinity of the moving object are detected by a corresponding radio receiver on the moving object. This captured sensor data is compared to previously captured and stored reference data received from a reference database. An association can then be made to determine one's current position based on previously measured signal strengths at reference positions. Previously measured signal strengths at previously known positions are analyzed and a similarity to the current measurement is determined to determine the position of the moving object.

A signal-strength-based localization based on the existing infrastructure is thus merged with a map-based or distance-based localization. The signal-strength-based localization is comparatively imprecise, while the map-based or distance-based localization enables a high level of accuracy, but can lead to ambiguities in the position estimate. By combining the two methods, a clear localization with high accuracy is achieved.

Compared to previous approaches, it is particularly advantageous that no new infrastructure has to be installed. In most environments, wireless communication technology is already installed anyway due to the connectivity needs of mobile robots or mobile devices. The approach according to the invention does not require the positions of the individual transmitters of this wireless communication technology to be known. Localization can be carried out by accessing a previously established reference database. This results in localization that can be implemented efficiently, even in dynamic environments.

In a preferred embodiment, the communication interface for transmitting the signal strength data and the final position estimate tion to the reference database. The reference database is therefore supplied with current information and updated to this extent. By storing an accurate and current position in conjunction with the signal strengths received at that position (fingerprint of a position), changes in the environment can be detected and taken into account. Constantly updated reference data is thus collected in the central reference database. In systems in which several moving objects are operated in parallel, in particular, a constantly updated and highly accurate reference database can be set up and maintained. In addition, no manual intervention is required when updating the reference database, so that the effort is kept to a minimum.

In a preferred embodiment, the first localization unit is designed to determine the first position estimate based on a two-dimensional scan matching algorithm. Various scan matching approaches are known in the prior art. In particular, an efficiently calculable two-dimensional approach can be used here, with localization being undertaken within a two-dimensional floor plan of the environment. A localization on several floors can then, for example, be mapped in a simple manner by additional information on the floor (also referred to as 2.5D localization). There are various algorithmic approaches for assigning detected contours to known contours (scan matching or contour comparison).

In a preferred embodiment, the distance interface is designed to receive the distance data from a lidar sensor. Preferably, the distance data can be received from a lidar sensor with a 360° field of view. With a lidar sensor, light signals are emitted and then received again after being reflected by objects in the vicinity. Distances to objects can be determined with comparatively high accuracy. In particular, a scanning approach can be used in which an environment is scanned two-dimensionally parallel to a subsurface plane with a rotating light beam. An efficiently realizable possibility for determining the distances to objects in the environment results.

In a preferred embodiment, the distance interface is designed to receive a two-dimensional contour. A two-dimensional contour is to be understood in particular as an image of the field of view arranged parallel to a background plane. All visible objects or all surfaces of the visible objects are recorded so that a kind of floor plan of a visible area can be generated. Such a two-dimensional contour can be efficiently processed algorithmically, for example based on scan-matching algorithms.

In a preferred embodiment, the signal strength interface is designed to receive signal strengths from access points of a wireless communication network, in particular a WiFi network. The access points are identified by their MAC addresses. In this respect, a WiFi fingerprinting approach is implemented. In most environments, for example in production halls or in office or administration buildings, WiFi networks are already available to enable wireless communication. These have various access points that act as transmitters that emit wireless signals whose signal strengths can be recorded. An existing infrastructure is used, so to speak, so that efficient installability and efficient operation are possible.

In a preferred embodiment, the second localization unit is designed to determine the second position estimate based on a least spring squares approach. A least spring squares approach is understood to be an approach in which distances in individual dimensions are each squarely evaluated as a metric for determining a multidimensional deviation. It is a standard approach that enables an efficiently calculable evaluation of distances. Accurate and efficient localization can be achieved.

In a preferred embodiment, the second localization unit is designed to determine the second position estimate based on an interpolation method. In particular, a linear or a model-based interpolation method is used. An interpolation is required in particular when an intermediate value between two known positions or two fingerprints is to be used. There are different approaches to achieve accurate results on the one hand and to ensure efficient calculability on the other.

In a preferred embodiment, the communication interface is designed to receive reference data with a list of positions and assigned signal strengths. The second localization unit is designed to determine the second position estimate as that position on the list at which the signal strength data have the smallest deviation from the signal strength data on the list. In particular, a list of positions and associated signal strengths can be received will. The signal strengths of different transmitters are assigned to a position in a so-called fingerprint. A list of fingerprints is received. A deviation of the current measurement from the fingerprints can then be made. The known position that has the smallest deviation is identified as the current position. Efficient calculability with high accuracy results.

In a preferred embodiment, the evaluation unit is designed to determine the final position estimate based on an estimation method. In particular, a particle filter or an extended Kalman filter can be used as an estimation method. An extended Kalman filter can, under certain circumstances, enable an optimal solution if measurement errors occur with normal distribution. In addition, an extended Kalman filter can be calculated comparatively efficiently. A particle filter can also provide good results for measurement errors of any statistical distribution, but is often more complex to calculate. Depending on the application scenario, an estimation method suitable for this application scenario can be selected. The result is a high level of accuracy when determining the final position estimate.

A moving object can in particular be a robotic vehicle in an industrial environment. However, a moving object can also be a pedestrian or another vehicle. A previously known environment is an environment of which a map is available. In particular, a floor plan of a building or of several buildings can be available. The device according to the invention is particularly suitable for use in environments that are at least partially inside buildings and in which satellite navigation systems therefore only function to a limited extent. A position of the moving object can be specified in particular in the form of coordinates in a predefined coordinate system relative to the map of the previously known environment. In particular, a two-dimensional position can be used. A position estimate preferably includes both an indication of the position (for example in the form of coordinates) and an indication of the accuracy or reliability of this position, for example in the form of a confidence value or a variance or standard deviation. A surrounding object can be both an object of the infrastructure, i.e. a wall, a door or similar, and an object that can move relative to the infrastructure, for example a box with production materials, another vehicle, etc.

• 1 eine schematische Darstellung eines erfindungsgemäßen Systems zum Lokalisieren beweglicher Objekte;
• 2 eine schematische Darstellung einer erfindungsgemäßen Vorrichtung zum Bestimmen einer Position eines beweglichen Objekts;
• 3 eine schematische Darstellung einer Lokalisierung basierend auf Abstandsdaten;
• 4 eine schematische Darstellung einer Lokalisierung basierend auf Signalstärkedaten;
• 5 eine schematische Darstellung einer finalen Positionsschätzung; und
• 6 eine schematische Darstellung eines erfindungsgemäßen Verfahrens.

The invention is described and explained in more detail below using a few selected exemplary embodiments in connection with the accompanying drawings. Show it:

• 1 a schematic representation of a system according to the invention for locating moving objects;
• 2 a schematic representation of a device according to the invention for determining a position of a moving object;
• 3 a schematic representation of a localization based on distance data;
• 4 a schematic representation of a localization based on signal strength data;
• 5 a schematic representation of a final position estimate; and
• 6 a schematic representation of a method according to the invention.

In the 1 A system 10 according to the invention for locating moving objects 12 in a previously known environment 14 is shown schematically. The system 10 comprises a device 16 for determining a position of a moving object 12, a distance sensor 18 which is arranged on the moving object 12, a reference database 20 and a plurality of transmitters 22a-22g. The illustration is to be understood as a top view of the surroundings 14 from a bird's eye view.

In the exemplary embodiment shown, the system 10 according to the invention is used in a production environment in which mobile agents are to be localized as moving objects 12 . The moving objects 12 can, for example, be units of a more or less autonomously operating driverless transport system. Such a transport system can be used, for example, to transport pallets or goods within the production environment.

The positions of the moving objects 12 should be made continuously detectable as far as possible. On the one hand, based on the positions, the mobile agents can navigate and, on the other hand, logistics planning or the like can also be carried out at a central location. The environment 14 is already known insofar as a floor plan or a map is available. The position of the moving object 12 is determined in particular within this map.

Surrounding objects 23 are located in the surrounding area. A surrounding object 23 can be, for example, a transport box, a production facility or a wall. One challenge lies in the changing environmental conditions For example, the arrangement of production facilities can be changed or materials can be parked in some places, which temporarily changes the topography of the area or the layout of the area. The localization should be as robust as possible against such changes. In addition, a localization is to be provided that requires the least possible intervention in the existing infrastructure.

The transmitters 22a-22g can in particular be access points (access points) of a WLAN network (in particular a WiFi network) which is operated in the environment for establishing wireless data communication. It is understood that other wireless communication approaches can also be used.

The reference database 20 can be designed in particular as a central database in the data communication network. In other words, a server in the network of the production environment can be used as reference database 20 . It goes without saying, however, that it is also possible to use a cloud database.

In the 2 a device 16 according to the invention for determining a position of a moving object within a previously known environment is shown schematically. The device includes a distance interface 24, a first localization unit 26, a signal strength interface 28, a communication interface 30, a second localization unit 32 and an evaluation unit 34. The individual units can be partially or fully implemented in software and/or hardware. In particular, the units can be designed as processors, processor modules or also as software for a processor. The device 16 can be designed in particular in the form of a control device of an autonomous or semi-autonomous vehicle or as software for a control device of an autonomous or semi-autonomous vehicle.

Distance data is received from a distance sensor via the distance interface 24 . The distance sensor can be a lidar or radar sensor, for example, which is arranged on the moving object. The distance sensor allows detection of distances between the moving object and surrounding objects in the area. Surrounding objects can be, for example, walls, production facilities, parked objects or even vehicles or people. Such objects are detected by the distance sensor within a field of view. In particular, objects can be detected in a 360° field of view. For this purpose, for example, a lidar sensor can be arranged on top of a vehicle in such a way that a 360° all-round view is made possible. In this case, the distances are in particular continuously recorded with a predetermined sampling rate. Because distances within the field of view are detected, a two-dimensional contour of the surrounding objects can be detected within the field of view.

A first position estimate is determined in the first localization unit 26 based on the distance data and based on a previously known map of the surroundings and a previous position estimate of the moving object. A scan-matching algorithm can be used here, for example, by means of which the detected contours within the field of view are compared with contours from the previously known map. A positioning is carried out, so to speak, by comparing the currently perceived sensor data with the map. The additional consideration of the previous position estimate can significantly reduce the computation effort and improve the accuracy of the position estimate. A position estimate is preferably understood to mean an indication of a mean value and a variance or a standard deviation in two dimensions.

Signal strength data is received via the signal strength interface 28 . In particular, the signal strengths of radio signals to various WiFi access points in the area are received or detected. For this purpose, the signal strength interface 28 can be connected to a WiFi module, for example. Usually, wireless communication systems with multiple access points allow detection of multiple access points and statement regarding the signal quality of the various access points. The signal strength data includes, in particular, so-called RSSI values from the various transmitters.

Reference data are received from the reference database via the communication interface 30 . These reference data are preferably several pairs of position information and signal strengths of different transmitters associated with this position information. A list of signal strengths of the transmitters that are received at this position is therefore transmitted for different positions.

If an accuracy of the first position estimate is below a threshold value, a second position estimate is determined in the second localization unit 32 . For this purpose, the received signal strength data, i.e. the current measurement data, are compared with the reference data. the This approach is referred to in particular as WiFi fingerprinting. For example, a deviation from the current signal strength data can be determined for each previously known position of the reference data and the signal strength values assigned to this previously known position. That previously known position of the reference data is then used as the second position estimate, which results in the smallest deviation between the signal strengths of the reference data and the received signal strength data. A least squares error approach can be used for this purpose, for example. It is also possible to use an interpolation method in which interpolation takes place between several previously known positions of the reference data.

A final position estimate is determined and output by the evaluation unit 34 . On the one hand, the first position estimate can be used as the final position estimate if the accuracy of the first position estimate is already sufficient and is above a threshold value. Either a previously known threshold value or a dynamically adapted threshold value can be used. If the accuracy of the first position estimate is below the threshold, the final position estimate is determined based on the second position estimate, the distance data, the previously known map of the environment, and the previous position of the moving object. An estimation method, in particular a particle filter or an extended Kalman filter, can be used for this purpose. A stochastic optimization method of this kind is used to determine one's own position as accurately as possible, based on assumptions or known properties of the measurement inaccuracies.

In the 3 , 4 and 5 the approach according to the invention for determining the position of the moving object 12 is illustrated schematically.

The 3 the horizon perceived by the moving object 12 or the distance sensor 18 arranged thereon. In particular, a contour 36 of the surroundings is perceived. If a previous position estimate is now known or the currently detected contour can be clearly assigned to a map section, the moving object 12 can be localized. In cases where the previous position estimate is inaccurate, for example because the moving object has been relocated while powered off and/or has been restarted, a localization may not be possible. On the one hand, mapping contours on large maps is computationally intensive. On the other hand, ambiguous situations can arise if the same contours are present at different points. In other words, a global position search using a scan-matching approach is possible, but often not unambiguous because contours are repeated. For example, similar machining centers can be present in a production hall. This effect is intensified by the fact that contours can change locally due to mobile objects (such as material containers).

If it is determined that the accuracy of the first position estimate, which can be determined by evaluating the distance data or the contours, is too low, a further position estimate is initiated. For this is in the 4 shown that the moving object 12 can receive radio signals from three transmitters 22a, 22b, 22e, with each transmitter (access point) being received with a different signal strength (cf. table on the right-hand side of the 4 for the measurement). To this extent, this signal strength data represents a measurement of the currently received signal strengths of the radio signals at the current position of the moving object 12 . In order to localize the movable object 12, the current measurement can then be compared with reference data from the reference database 20. In the exemplary embodiment shown, these reference data include a list 38 of positions and signal strengths assigned to these positions (left-hand side in FIG 4 ). A comparison of the current measurement with these reference data is therefore used to determine whether the moving object 12 is currently in a position that is close to one of the reference positions on the list 38 . For this purpose, for example, a least squares error approach can be used or an interpolation between several reference positions can be used. In this respect, the second position estimate is used for rough localization, which can then be further refined based on the data used for the first position estimate. In other words, all available data is merged if the first position estimate cannot be determined with sufficient accuracy.

It is particularly advantageous here if a plurality of moving objects each transmit their final position estimate and the signal strength data back to the reference database 20 . Like in the 5 shown, a constantly updated reference database can be achieved by communicating the current positions and the signal strengths received at these positions. In this respect, the signal strength database is continuously updated. By such a swarm approach, in which several moving objects 12 are in the same in regular operation move around, it is possible to react to changes in the infrastructure. The fingerprints of the reference database are continuously updated.

In the 6 a method according to the invention for determining a position of a moving object within a previously known environment is shown schematically. The method comprises steps of receiving S10 distance data, determining S12 a first position estimate, receiving S14 signal strength data, receiving S16 reference data, determining S18 a second position estimate and outputting S20 a final position estimate. Optionally, the method includes a step of transmitting S22 the signal strength data and the final position estimate to the reference database. In particular, the method can be embodied in software that is used on a processor of an autonomously operating vehicle. It goes without saying that the method can also be implemented as a smartphone app in order to be used in personal navigation.

The invention has been comprehensively described and explained with reference to the drawings and the description. The description and explanation are intended to be exemplary and not limiting. The invention is not limited to the disclosed embodiments. Other embodiments or variations will become apparent to those skilled in the art upon use of the present invention upon a study of the drawings, the disclosure, and the claims that follow.

In the claims, the words "comprising" and "having" do not exclude the presence of other elements or steps. The undefined article "a" or "an" does not exclude the presence of a plural. A single element or unit can perform the functions of several of the units recited in the claims. An element, unit, interface, device, and system may be partially or fully implemented in hardware and/or software. The mere naming of some measures in several different dependent patent claims should not be understood to mean that a combination of these measures cannot also be used to advantage. A computer program may be stored/distributed on a non-volatile medium, such as an optical memory or a solid state drive (SSD). A computer program may be distributed with and/or as part of hardware, for example via the Internet or via wired or wireless communication systems. Reference signs in the claims are not to be understood as restrictive.

Reference List

10

system

12

moving object

14

vicinity

16

contraption

18

distance sensor

20

reference database

22a-22g

Channel

23

environment object

24

distance interface

26

first localization unit

28

signal strength interface

30

communication interface

32

second localization unit

34

evaluation unit

36

contour

38

list

QUOTES INCLUDED IN DESCRIPTION

This list of documents cited by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Patent Literature Cited

• WO 2019/136019 A1 [0004]

Claims (13)

Device (16) for determining a position of a moving object (12) within a previously known environment (14), with: a distance interface (24) for receiving distance data from a distance sensor (18) with information on distances of the movable object (12) from surrounding objects in the environment (14) within a field of view of the distance sensor; a first localization unit (26) for determining a first position estimate of the moving object (12) based on the distance data, a previously known map of the surroundings (14) and a previous position estimate of the moving object (12); a signal strength interface (28) for receiving signal strength data with information on signal strengths of radio signals from a plurality of transmitters (22a-22g) in the vicinity (14); a communication interface (30) for receiving reference data from a reference database (20) with information on signal strengths of radio signals from the plurality of transmitters (22a-22g) at previously known positions in the area (14); a second locating unit (32) for determining a second position estimate of the moving object (12) based on a comparison of the signal strength data with the reference data when an accuracy of the first position estimate is below a threshold value; and an evaluation unit (34) for outputting a final position estimate, wherein the evaluation unit is designed to determine the final position estimate based on the second position estimate, the distance data, the previously known map of the surroundings (14) and the previous position of the moving object (12) if an accuracy of the first position estimate is below the threshold value, and is designed to output the first position estimate as a final position estimate if the accuracy of the first position estimate is above the threshold value. Device (16) after claim 1 , wherein the communication interface (30) is designed to transmit the signal strength data and the final position estimate to the reference database (20). Device (16) according to one of the preceding claims, wherein the first localization unit (26) is designed to determine the first position estimate based on a two-dimensional scan matching algorithm. Device (16) according to one of the preceding claims, wherein the distance interface (24) is designed to receive the distance data from a lidar sensor, preferably from a lidar sensor with a 360° field of view. Apparatus (16) according to any one of the preceding claims, wherein the distance interface (24) is adapted to receive a two-dimensional contour. Device (16) according to any one of the preceding claims, wherein the signal strength interface (28) is designed to receive signal strengths from access points of a wireless communication network, in particular a WiFi network; and the access points are identified by their MAC addresses. Device (16) according to one of the preceding claims, wherein the second localization unit (32) is designed to determine the second position estimate based on a least squares approach. Device (16) according to one of the preceding claims, wherein the second localization unit (32) is designed to determine the second position estimate based on an interpolation method, in particular a linear or a model-based interpolation method. Device (16) according to any one of the preceding claims, wherein the communication interface (30) is designed to receive reference data with a list of positions and associated signal strengths; and the second localization unit (32) is designed to determine the second position estimate as that position on the list at which the signal strength data have the smallest deviation from the signal strengths on the list. Device (16) according to one of the preceding claims, wherein the evaluation unit (34) is designed to determine the final position estimate based on an estimation method, in particular a particle filter or an extended Kalman filter. A system (10) for locating moving objects (12) within a known environment (14), comprising: an apparatus (16) according to any one of the preceding claims; a distance sensor (18) on the moving object (12) for receiving distances from surrounding objects; a reference database (20) for storing reference data and updating the reference data based on transmitted signal strength data and final position estimates; and a plurality of transmitters (22a-22g) for emitting radio signals. Method for determining a position of a moving object (12) within a previously known environment (14), with the steps: Receiving (S10) distance data from a distance sensor (18) with information on distances of the movable object (12) from surrounding objects in the area (14) within a field of view of the distance sensor; determining (S12) a first estimate of the position of the moving object (12) based on the distance data, a previously known map of the surroundings (14) and a previous estimate of the position of the moving object (12); receiving (S14) signal strength data with information on signal strengths of radio signals from a plurality of transmitters (22a-22g) in the area (14); receiving (S16) reference data from a reference database (20) with information on signal strengths of radio signals from the plurality of transmitters (22a-22g) at previously known positions in the area (14); determining (S18) a second position estimate of the moving object (12) based on a comparison of the signal strength data with the reference data if an accuracy of the first position estimate is below a threshold value; and Outputting (S20) a final position estimate, the final position estimate being determined based on the second position estimate, the distance data, the previously known map of the surroundings (14) and the previous position of the moving object (12) if an accuracy of the first position estimate is below the threshold is, and the first position estimate is output as the final position estimate if the accuracy of the first position estimate is above the threshold. Computer program product with program code for performing the steps of the method claim 12 , if the program code is run on a computer.

Images