WO2025253100A1

WO2025253100A1 - A spatial intelligence gathering system

Info

Publication number: WO2025253100A1
Application number: PCT/GB2025/051196
Authority: WO
Inventors: Edmund R HUNT; Timothy G PELHAM
Original assignee: University of Bristol
Current assignee: University of Bristol
Priority date: 2024-06-03
Filing date: 2025-06-02
Publication date: 2025-12-11
Anticipated expiration: 2026-12-03
Also published as: GB2641514A; GB202407869D0

Abstract

Aspects of the present invention relate to systems and methods of gathering intelligence, or data, about the physical layout and electromagnetic characteristics or electromagnetic behaviour of an environment. The method for mapping electromagnetic signals by one or more autonomous robots in an environment comprises mapping the physical layout of the environment and generating an electromagnetic model of the environment. The method further comprises detecting, by at least one of the one or more robots, an electromagnetic signal in the environment, and characterising the detected electromagnetic signal as a signal of interest. The method further comprises then directing the movement around the environment of at least one robot of the one or more robots based on the signal of interest, and identifying the source of the signal of interest in the environment.

Description

A SPATIAL INTELLIGENCE GATHERING SYSTEM

TECHNICAL FIELD

[0001] Embodiments described herein relate generally to systems and methods of gathering intelligence, or data, about the physical layout and electromagnetic characteristics or electromagnetic behaviour of an environment.

BACKGROUND

[0002] A wide variety of electrical devices can now be commonly found in a wide variety of different environments, both indoor and outdoor, such as computers, routers, electrical sockets and switches, lights, motors, and so on. Such electrical devices may be in continuous, near-continuous, periodic, or sporadic use, with each electrical device consequently generating and emitting electromagnetic fields or signals when in use. Such emitted electromagnetic signals then propagate through the environment in a complex manner. In some environments it can be necessary to monitor for new electrical devices, or electrical devices that are not expected or not wanted in that environment. Such monitoring often must be done manually, is time consuming, and needs to be repeated frequently, which can be disruptive to the normal functioning of that environment (e.g. an office or public location). In addition, the manner in which electromagnetic signals propagate through an environment can provide useful information on the environment itself, and can therefore be useful in a variety of scenarios, such as the structural health monitoring of buildings or other structures, or in the surveying of sites for archaeological, geological, or other engineering purposes. However, this too must often be done manually, can be time-consuming, and can require repeated or continuous monitoring, which can be disruptive.

SUMMARY OF INVENTION

[0003] The present application relates to the field of spatial intelligence gathering, for example physical mapping and/or electromagnetic mapping, of an environment.

[0004] In accordance with a first aspect of the invention, there is provided a method of mapping electromagnetic signals by one or more autonomous robots in an environment, the method comprising : mapping the physical layout of the environment, generating an electromagnetic model of the environment, detecting, by at least one of the one or more robots, an electromagnetic signal in the environment, characterising the detected i electromagnetic signal as a signal of interest, directing the movement around the environment of at least one robot of the one or more robots based on the signal of interest, and identifying the source of the signal of interest in the environment.

[0005] Any of the following may be applied to the above first aspect of the invention.

[0006] Optionally, the generating of the electromagnetic model of the environment is based on the mapping of the physical layout of the environment.

[0007] Optionally, the directing of the movement of at least one robot of the one or more robots based on the signal of interest comprises directing the at least one robot in a manner to identify the source of the signal of interest.

[0008] Optionally, the directing of at least one robot in a manner to identify the source of the signal of interest comprises the robot moving based on one or more of the characteristics of the signal of interest, the electromagnetic model of the environment, and/or the mapping of the physical layout of the environment.

[0009] Optionally, once the source of the signal of interest is identified, the directing of the movement of the at least one robot of the one or more robots is based upon further detected electromagnetic signals in the environment.

[0010] Optionally, the mapping of the physical layout of the environment is based on one or both of: at least one robot of the one or more robots collecting sensor data from the environment surrounding that robot, and/or retrieving data on the physical layout of the environment stored in a memory device.

[0011] Optionally, when the at least one robot of the one or more robots collects sensor data from the environment, one or more of the mapping of the physical layout of the environment, the generating of the electromagnetic model of the environment, or the directing of the movement around the environment of at least one robot of the one or more robots based on the signal of interest comprises directing the one or more robots in a closed loop manner.

[0012] Optionally, when the at least one robot of the one or more robots collects sensor data from the environment, that at least one robot employs simultaneous localization and mapping techniques to map the physical layout of the environment and/or to generate the electromagnetic model of the environment.

[0013] Optionally, the one or more robots is a plurality of robots, and the method further comprises: each robot of the plurality of robots communicating with another robot of the plurality of robots in order to carry out the steps of at least one of: mapping the physical layout of the environment; generating the electromagnetic model of the environment; directing the movement around the environment of at least one robot of the plurality of robots based on the signal of interest. [0014] Optionally, the characterising of the detected electromagnetic signal as a signal of interest comprises at least one of: determining that a characteristic of the detected electromagnetic signal is not an expected characteristic for electromagnetic signals for the environment, determining that the detected electromagnetic signal has an unexpected emission pattern over time, determining an approximate location of the source of the detected electromagnetic signal and determining that the determined approximate location is not an expected location for the detected electromagnetic signal, determining a device as being a possible source of the detected electromagnetic signal and determining that the device is not expected to be a source of the detected electromagnetic signal, and/or determining that the detected electromagnetic signal has not been detected before.

[0015] Optionally, the generating of the electromagnetic model of the environment comprises: emitting a first electromagnetic signal into the environment from an emitting location in the environment, the first electromagnetic signal having a set of emission characteristics, detecting, by at least one of the one or more robots, the first electromagnetic signal, the at least one of the one or more robots being at a respective receiving location within the environment, wherein the first electromagnetic signal detected by the at least one of the one or more robots has a corresponding set of received characteristics, modelling the propagation of the first electromagnetic signal based on the emitting location and the set of emission characteristics compared with each of the receiving locations and the corresponding sets of received characteristics.

[0016] Optionally, the emitting of the first electromagnetic signal is by one of: at least one robot of the one or more robots, a static emitter located at the emitting location in the environment, a non-autonomous mobile platform, or a mobile hand-held emitter at the emitting location in the environment.

[0017] Optionally, the emitting of the first electromagnetic signal is by at least one robot of the one or more robots, and wherein the detecting of the first electromagnetic signal comprises the same at least one robot of the one or more robots detecting the first electromagnetic signal.

[0018] Optionally, the one or more robots is a plurality of robots, and the directing of the movement of at least one robot comprises directing more than one of the plurality of robots based on the signal of interest, and the plurality of robots communicate with each other to identify the location of the source of the signal of interest within the environment.

[0019] Optionally, the one or more robots is a plurality of robots, and the determining that the detected electromagnetic signal is a signal of interest comprises a machine learning algorithm determining that the detected electromagnetic signal is a signal of interest, wherein training data for the machine learning algorithm comprises data corresponding to at least one of the plurality of robots emitting a simulated signal of interest from a location in the environment, and at least one of the plurality of robots detecting the simulated signal of interest and identifying the location as the source of the simulated signal of interest.

[0020] Optionally, the emitting location is a location in the environment at which a signal of interest might be expected to originate.

[0021] In accordance with a second aspect of the invention, there is provided a system for modelling electromagnetic signals in an environment, the system comprising : one or more autonomous robots, each robot of the one or more robots comprising : at least one sensor for collecting data such that the robot can navigate the environment, at least one sensor for collecting data relating to electromagnetic signals in the environment, and a communications module for transmitting the collected sensor data, wherein the system is configured to carry out any of the methods of the first aspect of the invention. [0022] The following may be applied to the above second aspect of the invention.

[0023] Optionally, one or more of the one or more robots further comprises a spatial intelligence node, the spatial intelligence node comprising: at least one sensor for localising the one or more of the one or more robots in the environment, and at least one antenna for emitting and receiving electromagnetic signals.

[0024] In accordance with a third aspect of the invention, there is provided a non- transitory memory storing one or more programs, which, when executed by one or more processors of a device, cause the device to perform any of the methods of the first aspect of the invention.

[0025] Within the scope of this application it is expressly intended that the various aspects, embodiments, examples and alternatives set out in the preceding paragraphs, in the claims and/or in the following description and drawings, and in particular the individual features thereof, may be taken independently or in any combination. That is, all embodiments and/or features of any embodiment can be combined in any way and/or combination, unless such features are incompatible. The applicant reserves the right to change any originally filed claim or file any new claim accordingly, including the right to amend any originally filed claim to depend from and/or incorporate any feature of any other claim although not originally claimed in that manner.

FIGURES

[0026] In the following, embodiments will be described with reference to the drawings in which : i [0027] Fig ure la shows an example scenario where three robots released into an environment generate a map of the physical layout of the environment, according to some embodiments.

[0028] Fig ure lb shows an example scenario where three robots move around the environment to generate or populate an electromagnetic model of the environment, according to some embodiments.

[0029] Fig ure 2a shows an example scenario where three robots move around the environment to monitor the environment and to identify electromagnetic (EM) signals of interest, according to some embodiments.

[0030] Fig ure 2b shows an example scenario where three robots are directed around the environment 100 based on an identified signal of interest, according to some embodiments.

[0031] Fig ure 3 shows a flowchart of a method for the spatial intelligence gathering system to gather intelligence on an environment, according to some embodiments.

[0032] Fig ure 4 shows an example scenario where three robots are directed around the environment and actively map the electromagnetic behaviour of the environment by emitting electromagnetic signals, according to some embodiments.

[0033] Fig ure 5 shows a flowchart of a method for the spatial intelligence gathering system to gather intelligence on an environment by actively mapping the electromagnetic behaviour of the environment, according to some embodiments.

[0034] Fig ure 6 shows a representation of the inputs and outputs for an example control algorithm, according to some embodiments.

DETAILED DESCRIPTION

[0035] The present application broadly relates to a spatial intelligence gathering system for use in a given environment, where a robotic system of the spatial intelligence gathering system is used to map the electromagnetic-physical environment and to identify signals of interest in that environment using said mapping.

[0036] Throughout the present application, the term "spatial intelligence node" is used to mean a suite of sensors used together as a "node" to gather information on the spatial surroundings around the sensor suite. In the embodiments of the present application, spatial intelligence nodes are installed on one or more robots, allowing for spatial intelligence to be gathered as the robots move around an environment.

[0037] Throughout the present application, references to "electromagnetic signals" will be understood as referring to any signal falling across the electromagnetic spectrum, and any such electromagnetic signals as including "electromagnetic waves" or "electromagnetic fields".

[0038] Throughout the present application, references to the propagation of electromagnetic signals through an environment will be understood as including the scattering of such signals off of objects and surfaces within that environment and the transmission or absorption of such signals by objects and surfaces within that environment. As such, this includes the reflection, diffraction, or otherwise scattering of, as well as the refraction, absorption, or otherwise transmitting of, such electromagnetic signals as they interact with objects or surfaces within the environment. This also encompasses any associated changes to an electromagnetic signal arising from such interactions with the environment during propagation (e.g. changes in polarisation, frequency shifts or other frequency changes, or any other change in electromagnetic signal characteristics).

[0039] It will be understood that any references to electronic devices or electronic components found the present application include any electrical device or component as well as any electromechanical devices and components.

[0040] It will also be understood that any references included herein to electromagnetic signals present in an environment includes electromagnetic signals generated by electronic devices or electronic components within the environment as a consequence of current flowing in those electronic devices or electronic components. As such, any reference to such electronic devices or electronic components being "in use" include scenarios in which a current is flowing through such electronic devices or electronic components and therefore such electromagnetic signals (or waves or fields) are consequently generated.

[0041] Fig ures la to 2b show an example approach to spatial intelligence gathering according to some embodiments of the present disclosure. In particular, Figures la and lb show an approach in which one or more robots 200a to 200n of a spatial intelligence gathering system 10 are equipped with a spatial intelligence node 50 and used to generate a map of the physical layout 110 of the environment, as well as populate an electromagnetic model 120 of the environment 100. The one or more robots 200a to 200n of the spatial intelligence gathering system 10 may also be used to categorise or identify detected signals as signals of interest 70 in an environment 100 based on the physical map 110 of the layout of that environment 100 and/or the electromagnetic model 120 of that environment 100, and may employ the spatial intelligence node 50 included in the one or more robots to do so, as shown in Figure 2a. The sources in the environment 100 of such signals of interest 70 can then be identified using the one or more robots 200a to 200n_z as shown in Figure 2b thereby providing i intelligence concerning the current spatial and electromagnetic state of the environment 100.

[0042] In Figure la, one or more robots 200a to 200n are released or deployed into an environment 100. In some embodiments, and in the example scenario shown in Figure la, three robots 200a, 200b, and 200c are released into the environment 100.

[0043] It will be understood that in any appropriate embodiment discussed herein, the one or more robots (e.g. the three robots 200a to 200c of the example scenarios shown in Figures la to 2b) may be autonomous, and may take the form of, for example, ground-based autonomous robots (e.g. wheeled or tracked robots), autonomous aerial drones, or waterborne drones, or any suitable combination based on the particular environment 100 and scenario in question.

[0044] The one or more robots 200a to 200n are each equipped with at least one antenna 220 for sending and receiving signals. Here, the signals emitted and received by the at least one antenna 220 of each of the one or more robots 200a to 200n may be to and from other robots of the one or more robots 200a to 200n_z to and from a server 250 of the spatial intelligence gathering system 10, or for other purposes discussed herein. In some embodiments, the emission and reception of signals by the at least one antenna 220 may be directed in a chosen direction, for example by using a horn antenna, beam steering, beam forming, or other suitable approach (e.g. where the at least one antenna 220 is an antenna array).

[0045] In some embodiments, the at least one antenna 220 may be mounted on one or more actuators or motorised arms of a robot, allowing the at least one antenna 220 to be redirected to focus on, for example, a particular area, region, or surface in the environment 100 (e.g. a wall). As a result, in such embodiments the at least one antenna 220 is able to pan and/or tilt to be directed (or "scan") left, right, up, or down to be directed in a chosen direction within the environment 100.

[0046] In some embodiments, the server 250 may be an off-site server of the spatial intelligence gathering system 10, and may communicate with the one or more robots 200a to 200n over an internet connection. In some embodiments, the server 250 may be a cloud server of the spatial intelligence gathering system 10, and may communicate with the one or more robots 200a to 200n over an internet connection. In some embodiments, the server 250 may be a computing device, such as a computer, laptop, or mobile device, that may be present on-site (e.g. in or close to the environment 100) and communicate wirelessly directly to the one or more robots 200a to 200n (e.g. by a direct wireless connection or over a local wireless network). In some embodiments, the server 250 may communicate with the robots periodically via a direct connection (e.g. a docking station), where data is communicated to the server 250 (or received from the server 250) in periodic batches (e.g. when one or more of the robots 200a to 200n is docked). Such communication may be achieved using a communications module 240 forming part of the spatial intelligence node 50 on each robot of the one or more robots 200a to 200n.

[0047] In some embodiments, the one or more robots 200a to 200n may additionally include one or more sensors 230, such as a camera 230a, a 2D and/or 3D lidar sensor 230b, an infrared sensor 230c, a thermal sensor 230d, a depth camera 230e, a hyperspectral camera 230f, and/or a sonar sensor 230g. It will be appreciated that other sensors 230 may be employed on the one or more robots 200a to 200n.

[0048] In some embodiments, the one or more sensors 230 may be mounted on one or more actuators or motorised arms of a robot, allowing the one or more sensors 230 to be redirected to focus on, for example, a particular area, region, or surface in the environment 100 (e.g. a wall). As a result, in such embodiments the one or more sensors 230 are able to pan and/or tilt to be directed (or "scan") left, right, up, or down to reach up or into a constricted space in the environment 100. This may be achieved whilst avoiding physical contact with a given object or surface in the environment 100, or may be achieved in order to interface with such an object or surface in the environment 100.

[0049] It will also be appreciated that, in embodiments where multiple robots are employed, each of the robots 200a to 200n may include the same or different sensors to the other robots of the robots 200a to 200n. Here, sensors may be chosen for the one or more robots 200a to 200n that are most appropriate to the environment 100 in question. As such, some or all of the robots of the one or more robots 200a to 200n may be equipped with such a spatial intelligence node 50, and the spatial intelligence node 50 of each robot need not be comprised of the same sensor systems as the spatial intelligence nodes 50 of other robots of the one or more robots 200a to 200n.

[0050] The one or more sensors 230 form part of the spatial intelligence node 50 included on each of the one or more robots 200a to 200n.

[0051] Returning to the embodiment of Figure la, robots 200a, 200b, and 200c released into the environment 100 may initially map the physical layout 110 of the environment 100 by autonomously moving around the environment 100 and collecting spatial sensor data using the onboard one or more sensors 230. For example, robots 200a, 200b, and 200c may record spatial sensor data from one or more of an onboard camera 230a, an onboard lidar sensor 230b, an onboard infrared sensor 230c, an onboard thermal sensor 230d, an onboard depth camera 230e, an onboard hyperspectral camera 230f, and/or an onboard sonar sensor 230g of the spatial intelligence node 50 included on one or more of the robots. As such, it will be understood that spatial sensor data relates to any i data regarding the physical layout and physical characteristics of the environment 100, and any objects located in the environment 100 (e.g. furniture, doors, equipment, light switches, power sockets, other wall fixings etc). In Figure la, the mapping of the physical layout 110 of the environment 100 by the robots 200a, 200b, and 200c, using the one or more sensors 230 of the spatial intelligence node 50 on each robot, is shown by a dotted segment attached to each robot 200a, 200b, and 200c.

[0052] In embodiments where multiple robots are employed, the robots 200a to 200n may communicate recorded spatial sensor data with each other (for example using their onboard antennas 220) in order to map the physical layout 110 of the environment 100. The robots 200a to 200n may alternatively, or in addition, communicate recorded spatial sensor data with the server 250 (when included) in order to map the physical layout 110 of the environment 100. Such communication may be achieved using a communications module 240 forming part of the spatial intelligence node 50 on each robot of the one or more robots 200a to 200n.

[0053] Each robot of the one or more robots 200a to 200n may generate a map of the physical layout 110 of the environment 100 based on the recorded spatial sensor data. The map of the physical layout 110 of the environment 100 may take the form of, for example, a 3D model or a point cloud. The map generated by each of the one or more robots 200a to 200n may be generated dynamically, by updating the map as additional spatial sensor data is recorded as each of the one or more robots 200a to 200n moves around the environment 100.

[0054] In some embodiments, the one or more robots 200a to 200n may generate the map of the physical layout 110 of the environment 100 by employing simultaneous localization and mapping (SLAM) techniques, where each robot localises its position within that map as it is generated. In some embodiments, this may be a form of visual SLAM (for example, when the environment 100 is an indoor environment). Alternatively, or additionally, any other suitable technique may be used (e.g. using a global navigation satellite system, GNSS, or by using fiducial markers, ultrasound beacons, Bluetooth low energy beacons, etc.).

[0055] In some embodiments, the one or more robots 200a to 200n may move autonomously around the environment in a closed-loop manner, where the movement of the one or more robots 200a to 200n is directed based on the spatial sensor data as it is gathered (either by the robot that is collecting the spatial sensor data, or by other robots of the one or more robots gathering spatial sensor data). In the example scenario of Figure la, the movement of each of the robots 200a, 200b, and 200c around the environment 100 is shown by an arrow. [0056] Each robot of the one or more robots 200a to 200n may communicate with other robots of the one or robots 200a to 200n (e.g. using the communications module 240 of the spatial intelligence node 50 of that robot), and/or with the server 250 (where included), to share the map of the physical layout 110 of the environment 100 with the other robots of the one or more robots 200a to 200n_z and/or with the server 250 (where included), as it is generated by that robot. This allows for a map of the physical layout 110 of the environment 100 to be generated and updated as the robots move autonomously and shared throughout the spatial intelligence gathering system 10. In this regard, the map of the physical layout 110 of the environment 100 may be updated dynamically and in real time as the robots move autonomously, or may be updated periodically (e.g. only when the robots come within communications range of each other or the server 250, or when the robots return to a docking station).

[0057] In some embodiments, at least one robot of the one or more robots 200a to 200n may only share the spatial sensor data recorded by that at least one robot with the other robots of the one or more robots 200a to 200n_z and the map of the physical layout 110 of the environment 100 may be generated by only some of the one or more robots 200a to 200n (e.g. only by a single robot that receives all the spatial sensor data recorded by the other robots). This reduces the computational requirements for some of the one or more robots 200a to 200n_z thereby reducing power consumption, size, and/or weight for those robots of the one or more robots 200a to 200n.

[0058] In some embodiments, the server 250 may receive all the spatial sensor data recorded by the one or more robots 200a to 200n, and may generate the map of the physical layout 110 of the environment 100. With this approach, the computational requirements for each robot of the one or more robots 200a to 200n is reduced, thereby reducing the power consumption of each robot (amongst other benefits).

[0059] In some of embodiments, including those discussed above, the dynamically generated map of the physical layout 110 of the environment 100 (e.g. when generated by each robot, only some robots, or by the server 250) may be used to inform the movement of each robot of the one or more robots 200a to 200n around the environment 100 as the map is continually generated (e.g. in a closed loop manner).

[0060] For example, when the dynamically generated map of the physical layout 110 of the environment 100 shows that a part of the environment 100 has been sufficiently mapped, the one or more robots 200a to 200n may communicate with each other (e.g. using the communications module 240 of the spatial intelligence node 50 of each robot) to direct their autonomous movements such that that part of the environment 100 is not re-mapped unnecessarily. As a further example, each robot of the plurality of robots 200a to 200n may store an up-to-date copy of the generated map of the physical layout 110 of the environment 100 and will autonomously move through the environment 100 in a manner to minimise, or avoid entirely, remapping parts of the environment 100 unnecessarily that have already been mapped by itself or by another robot of the one or more robots 200a to 200n.

[0061] As an additional example, in embodiments that include the server 250, the server 250 may direct the movement of the one or more robots 200a to 200n in order to ensure that any unnecessary remapping of the environment 100 is minimised or avoided entirely.

[0062] In some embodiments, the dynamically generated map of the physical layout 110 of the environment 100 may indicate that an area of the environment 100 has been insufficiently mapped. As such, the one or more robots 200a to 200n (or, where included, the server 250) may identify such insufficiently mapped areas and may communicate with each other (e.g. using the communications module 240 of the spatial intelligence node 50 of each robot) to direct the movement of one robot of the one or more robots 200a to 200n to remap that area of the environment 100 as needed. In some embodiments, certain areas of the generated map of the physical layout 110 of the environment 100 may be selected (e.g. by the system 10) for more detailed mapping (for example, certain closed spaces in the environment 100, such as cavities, or certain fixtures or fittings). The one or more robots 200a to 200n may then map those areas of the environment 100 to the sufficient level of detail as needed.

[0063] In some embodiments, the one or more robots 200a to 200n may receive a map of the physical layout 110 of the environment 100 that has previously been generated (for example, building CAD or Digital Surface/Terrain Models of the environment in question). For example, where the spatial intelligence gathering system 10 has previously generated a map of the physical layout 110 of the environment 100, or where another system has previously done so, the one or more robots 200a to 200n may receive and store that previously generated map of the physical layout 110 of the environment 100.

[0064] In such scenarios, it may be necessary only for the one or more robots 200a to 200n to update the received previously generated map of the physical layout 110 of the environment 100 to reflect the current configuration of the environment 100 (e.g. where objects have since been moved or introduced, such as moved furniture or new electrical devices). In such scenarios, it may not be necessary for the one or more robots 200a to 200n to carry out any additional mapping of the physical layout 110 of the environment 100 (other than to account for such changes in the current configuration of the environment 100, if any) beyond what is included in the received previously generated map of the physical layout 110 of the environment 100. li [0065] The generated map of the physical layout 110 of the environment 100 may take the form of a two-dimensional or three-dimensional representation of the environment 100 (for example by scanning the environment using a 2D or 3D lidar sensor), or may take the form of data that represents a three-dimensional representation of the environment 100. As discussed above, the map of the physical layout 110 of the environment 100 may take the form of, for example, a 3D model or a point cloud. The generated map of the physical layout 110 of the environment 100 may include rooms, hallways, and open spaces. The generated map of the physical layout 110 of the environment 100 may also include objects in the environment 100, such as tables, chairs, or other furniture. The generated map of the physical layout 110 of the environment 100 may include openings and apertures in walls, such as doorways (with or without doors), windows, or skylights, as well as access points to narrow spaces such as above ceiling plenum spaces, interstitial spaces, attics, storage spaces (e.g. cupboards, wardrobes, etc.), wall cavities, or air conditioning ducts. In the example scenario of Figure la, doorways 105a, windows 105b, desks 105c, and cupboards 105d are identified, but it will be understood that a variety of other physical aspects of the environment 100 may be identified. The generated map of the physical layout 110 of the environment 100 may include areas where the environment 100 is outdoors, such as a courtyard or garden. In some embodiments, the generated map of the physical layout 110 of the environment 100 may comprise an entirely outdoor environment.

[0066] The generated map of the physical layout 110 of the environment 100 may additionally include electrical devices and electrical components located within the environment 100. For example, the generated map of the physical layout 110 of the environment 100 may include the locations of computers (such as desktops or laptops), electrical outlets (such as power sockets, network plugs, light fittings, and light switches), as well as the location of any other electrical components or wiring (such as wireless hubs, electrical wires, desk-based phones, or wall clocks).

[0067] In some embodiments, the one or more robots 200a to 200n (and/or, where included, the server 250) may identify such electrical devices and electrical components located in the environment 100 as part of generating the map of the physical layout 110 of the environment 100, or may identify such electrical devices and electrical components located in the environment 100 after generating the map of the physical layout 110 of the environment 100. For example, the one or more robots 200a to 200n (and/or, where included, the server 250) may identify the location of desktop computers, laptops, light switches, power sockets, wireless hubs, network plugs, and electrical wires within the environment 100 as part of the map of the physical layout 110 of the environment 100. In the example scenario of Figure la, such identified electrical devices and electrical components include desktop computers and screens 105e, telephones 105f, electrical wall sockets 105g, and electrical light switches 105h, although it will be understood that a variety of other types of electrical devices and electrical components may be identified.

[0068] In some embodiments, a machine learning algorithm or technique may be employed to identify the electrical devices and electrical components located in the environment 100, and update the map of the physical layout 110 of the environment 100 accordingly. For example, a machine learning algorithm may be utilised to identify desktop computers, power sockets, WiFi hubs or routers, light switches, or other electrical devices or components present in the environment 100, based on the spatial sensor data from the one or more sensors 230 of the spatial intelligence node 50 on each of the one or more robots 200a to 200n.

[0069] In some embodiments, the same or a different machine learning algorithm or technique may be employed to identify the likely material properties of objects or structures in the environment 100, and may update the map of the physical layout 110 of the environment 100 accordingly. For example, such a machine learning algorithm may identify a wall as being composed of concrete or brick, a table as being composed of wood or metal, a chair as being composed of plastic or upholstered materials, or the ground as being composed of grass, asphalt, or concrete. Such a machine learning algorithm or technique may make such an identification based on the spatial sensor data from the one or more sensors 230 of the spatial intelligence node 50 on each of the one or more robots 200a to 200n. Alternatively, or in addition, the map of the physical layout 110 of the environment 100 may be updated by a user to include the material properties of objects or structures in the environment 100.

[0070] The material properties of objects or structures in the environment 100 may then be utilised in the generation or population of the electromagnetic model 120 of the environment 100, since such materials may affect the propagation of electromagnetic signals through the environment 100 (i.e. such materials may result in different propagation or scattering behaviour of EM signals).

[0071] In some embodiments, the generating of the physical layout 110 of the environment 100 may include recording spatial sensor data at different times to assess how the environment 100 changes over a period of time (e.g. forming a "pattern of life" for the environment 100). For example, the one or more robots 200a to 200n may generate the physical layout 110 of the environment 100 over multiple sessions at different times, such as once during a working weekday (where the environment 100 is, for example, an office), and once during a weekend. Other time periods may include, for example, different times of day, different weeks in a typical month, or different times of year.

[0072] Fig ure lb shows the generation of an electromagnetic model 120 of the environment 100 by the one or more robots 200a to 200n in the environment 100. To continue the example scenario of Figure la above, Figure lb shows three robots 200a, 200b, and 200c carrying out the generation of an electromagnetic model 120 of the environment 100.

[0073] As discussed above, the one or more robots include a spatial intelligence node 50 that comprises the one or more sensors 230. The one or more sensors 230 may additionally include one or more electromagnetic (EM) sensors 230h, where the one or more EM sensors 230h are configured to detect EM signals (e.g. radio signals) in the environment 100. In some embodiments, the one or more EM sensors 230h may include one or more antennas.

[0074] In some embodiments, the one or more EM sensors 230h may be mounted on one or more actuators or motorised arms of a robot, allowing the one or more EM sensors 230h to be redirected to focus on, for example, a particular area, region, or surface in the environment 100 (e.g. a wall). As a result, in such embodiments the one or more EM sensors 230h are able to pan and/or tilt to be directed (or "scan") left, right, up, or down to reach up or into a constricted space in the environment 100. This may be achieved whilst avoiding physical contact with a given object or surface in the environment 100, or may be achieved in order to interface with such an object or surface in the environment 100. For example, the one or more EM sensors 230h may then be directed downwards towards the ground for buried infrastructure survey, or may be directed to physically contact or interface with the surface of an object in the environment 100 such as a wall or the ground.

[0075] It will be understood that, in embodiments where the at least one antenna 220, the one or more sensors 230, and/or the one or more EM sensors 230h are installed on one or more actuators or motorised arms, the at least one antenna 220, the one or more sensors 230, and/or the one or more EM sensors 230h may be installed on the same or a different one or more actuators or motorised arms of the robot in question. It will also be understood that only some or all of the one or more robots 200a to 200n may include such one or more actuators or motorised arms, and that the one or more actuators or motorised arms on a given robot may include the same or different antenna, sensor, or sensors to others of the one or more robots 200a to 200n.

[0076] Here, electromagnetic (EM) data collected by the one or more EM sensors 230h of the spatial intelligence node 50 on each robot (and in some embodiments in combination with sensor data from other sensors of the sensors 230, and/or in combination with the generated map of the physical layout 110 of the environment 100), is then used to generate an electromagnetic model 120 of the environment 100.

[0077] That is to say, the spatial intelligence node 50 included on the one or more robots 200a to 200n is able to use the one or more sensors 230 to generate an electromagnetic model 120 of the environment 100 (in addition to, in some embodiments, generating the map of the physical layout 110 of the environment 100). [0078] In some embodiments, the electromagnetic model 120 of the environment 100 may take the form of a channel model. Here, the term "channel model" is understood to mean a representation of the effects of a communication channel (e.g. a medium or environment) through which wireless signals are propagated.

[0079] For instance, the channel model included in the electromagnetic model 120 of the environment 100 may be able to predict the power loss, changes in phase, and/or changes in the polarisation incurred by a signal as it travels through the environment 100. As a further example, ray tracing may be used with in combination with the map of the physical layout 110 of the environment 100 to assess the presence of scatterers and/or transmitters in the environment 100 that may reflect, or transmit, transmitted electromagnetic signals in the environment 100 to a receiver in the environment 100. Where objects in the environment 100 may be moving, this may also include predicting changes in the frequency of a signal as it travels through the environment 100.

[0080] In some embodiments, multiple channel measurements may be taken for a single timestep (e.g. time period, such over a second), where the one or more robots 200a to 200n are equipped with a multichannel radio sensor. Such measurements taken over multiple channels may then be used to infer three-dimensional information about the environment 100 and the electromagnetic behaviour with each timestep, and create angle of arrival spectrums (i.e. azimuth/elevation) for detected EM signals.

[0081] Each robot of the one or more robots 200a to 200n moves autonomously around the environment 100 to generate the electromagnetic model 120 of the environment 100. Here, the principles of the movement of the robots may be the same as that discussed above with regard to the generation of the map of the physical layout 110 of the environment 100.

[0082] In some embodiments, the electromagnetic model 120 of the environment 100 is generated using SLAM techniques, or any other suitable technique, in the same manner as for that discussed above with regard to the generation of the map of the physical layout 110 of the environment 100.

[0083] Each robot of the one or more robots 200a to 200n may communicate with other robots of the one or robots 200a to 200n_z and/or with the server 250 (where included), to share the electromagnetic model 120 of the environment 100 with the other robots of the one or more robots 200a to 200n_z and/or with the server 250 (where included), as it is generated (e.g. dynamically) by that robot. This may be achieved using the communications module 240 of the spatial intelligence node 50 of each robot. This allows for the electromagnetic model 120 the environment 100 to be generated and updated as the robots move autonomously and shared throughout the spatial intelligence gathering system 10. In this regard, the electromagnetic model 120 of the environment 100 may be updated dynamically and in real time as the robots move autonomously, or may be updated periodically (e.g. only when the robots come within communications range of each other or the server 250, or when the robots return to a docking station).

[0084] In some embodiments, at least one robot of the one or robots 200a to 200n may only share the EM data recorded by that at least one robot with other robots of the one or more robots 200a to 200n_z and the electromagnetic model 120 of the environment 100 may be generated by only some of the one or more robots 200a to 200n (e.g. only by a single robot that receives all the EM data recorded by the other robots). This reduces the computational requirements for some of the one or more robots 200a to 200n_z thereby reducing power consumption for those robots of the one or more robots 200a to 200n. This may also allow for reduced weight, size, and/or cost of some of the one or more robots 200a to 200n. A reduced weight and/or size of a given robot allows such a robot to enter more hard-to-reach areas of an environment (e.g. inside air ducts, access ducts, or when operating underwater).

[0085] In some embodiments, the server 250 may receive all the EM data recorded by the one or more robots 200a to 200n_z and may generate the electromagnetic model 120 of the environment 100. With this approach, the computational requirements for each robot of the one or more robots 200a to 200n is reduced, thereby reducing power usage of each robot.

[0086] In some of embodiments, including those discussed above, the dynamically generated electromagnetic model 120 of the environment 100 (e.g. when generated by each robot, only some robots, or by the server 250) may be used to inform the movement of each robot of the one or more robots 200a to 200n around the environment 100 as the electromagnetic model 120 is continually generated or populated (e.g. in a closed loop manner).

[0087] In some embodiments, generating or populating the electromagnetic model 120 may include employing the map of the physical layout of the environment 100. For example, the map of the physical layout 110 of the environment 100 may be used in directing the one or more robots 200a to 200n more efficiently around the environment ii 100 to ensure that the electromagnetic model 120 is more detailed and accurate for the environment 100 in question. As a further example, the physical layout 110 of the environment 100 may indicate that a given surface (e.g. a wall) is made from a certain material with given reflection, scattering, and/or absorbing properties for EM signals, which may be used in the generation or population of the electromagnetic model 120, thereby improving the accuracy of the model 120. Other spatial sensor data recorded as part of the generation of the physical layout 110 of the environment 100 may be used in an analogous manner in the generation or population of the electromagnetic model 120.

[0088] In some embodiments, the electromagnetic model 120 of the environment 100 is generated at the same time as the generation of the map of the physical layout 110 of the environment 100. That is to say, the one or more robots 200a to 200n may move autonomously around the environment 100 in the manner discussed above, generating both the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100 simultaneously (e.g. using a SLAM approach, or other suitable approach such as those discussed above).

[0089] In some embodiments where the map of the physical layout 110 of the environment 100 has been previously generated (e.g. by an earlier deployment of the spatial intelligence gathering system 10, or by the earlier use of another system), the one or more robots 200a to 200n may move autonomously around the environment 100 recording EM data, and that EM data may be combined with the earlier map of the physical layout 110 of the environment 100 to generate the electromagnetic model 120 of the environment 100.

[0090] In some embodiments, the generation of the electromagnetic model 120 of the environment 100 may be done dynamically, as the one or more robots 200a to 200n move autonomously around the environment 100 recording EM data. In some embodiments, the generation of the electromagnetic model 120 of the environment 100 may be done after the one or more robots 200a to 200n have completed moving around the environment 100 and the EM data has been recorded.

[0091] In some embodiments, the one or more robots 200a to 200n may communicate with each other (e.g. using the communications module 240 of the spatial intelligence node 50 of each robot) to direct their autonomous movements such that the EM data for parts of the environment 100 are not re-recorded unnecessarily. For example, each robot of the plurality of robots 200a to 200n may store an up-to-date copy of the generated electromagnetic model 120 of the environment 100 and will autonomously move through the environment 100 in a manner to minimise, or avoid entirely, ii recording EM data from parts of the environment 100 that have already had EM data recorded by itself or by another robot of the one or more robots 200a to 200n.

[0092] As a further example, in embodiments that include the server 250, the server 250 may direct the movement of the one or more robots 200a to 200n in order to ensure that any unnecessary re-recording of EM data in the environment 100 is minimised or avoided entirely.

[0093] In some embodiments, the generated electromagnetic model 120 of the environment 100 may indicate that an area of the environment 100 requires further EM data to be recorded. As such, the one or more robots 200a to 200n (or, where included, the server 250) may identify such areas and may communicate with each other (e.g. using the communication module 240) to direct the movement of one robot of the one or more robots 200a to 200n to record further EM data in that area of the environment 100, and update the electromagnetic model 120 of the environment 100 as needed.

[0094] Returning to the example scenario of Figure lb, the robots 200a, 200b, and 200c move around the environment 100 and detect electrical signals emitted from the various electrical devices and components in the environment 100. Here, EM signals are emitted by each of the desktop computers and screens 105e, telephones 105f, electrical wall sockets 105g, and electrical light switches 105h in the environment 100 (shown in Figure lb as curved lines emitting from each of these electrical devices and electrical components). These EM signals then propagate through the environment 100, (for example by reflecting, refracting, and scattering off, or by being transmitted through, the various surfaces, in the environment 100) (shown in Figure lb by dotted curved lines). The robots 200a, 200b, and 200c detect these EM signals and utilise the resulting EM data to generate or populate the electromagnetic model 120 of the environment 100.

[0095] The electromagnetic model 120 of the environment 100 may then be used in combination with the map of the physical layout 110 of the environment 100 to predict the electromagnetic (EM) behaviour of electrical devices and electrical components identified in the environment 100. In some embodiments, the one or more robots 200a to 200n (or, where included, the server 250) may access a library or catalogue of typical EM signatures for a variety of different electrical devices and electrical components as part of predicting such EM behaviour. For example, this may include information such as typical frequency ranges and signal strength ranges of emitted signals for a given type of electrical device or electrical component.

[0096] For example, where a desktop computer has been identified in the map of the physical layout 110 of the environment 100, the electromagnetic model 120 may be used to predict how EM signals emitted from the desktop computer when in use might propagate through the environment 100. As a further example, where a doorway has been identified in the map of the physical layout 110 of the environment 100, the electromagnetic model 120 may be used to predict how EM signals present in the environment will propagate through that doorway into other parts of the environment 100. As an additional example, where a wall, floor, or ceiling has been identified in the map of the physical layout 110 of the environment 100, the electromagnetic model 120 may be used to predict how EM signals present in the environment 100 will scatter from that wall, floor, or ceiling, and/or propagate through that wall, floor, or ceiling, into other parts of the environment 100.

[0097] After the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100 have been generated, the one or more robots 200a to 200n can be used to monitor the environment 100 for EM signals present in the environment 100. Here, the spatial intelligence gathering system 10 may identify, characterise, or categorise some detected EM signals as being "signals of interest" based on the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100.

[0098] To continue the example scenario of Figures la and lb above, Figure 2a shows three robots 200a to 200c being released (or continuing to move around) the environment 100 to monitor the environment 100, and to identify electromagnetic (EM) signals of interest 70.

[0099] In some embodiments, the spatial intelligence gathering system 10 may carry out the monitoring of the environment 100 for EM signals of interest immediately after having generated the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100, or may carry out the monitoring of the environment 100 for electromagnetic EM signals of interest at a later time (e.g. monitoring the environment periodically, or for extended periods).

[0100] In some embodiments, the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100 may have been generated by a system separate to the spatial intelligence gathering system 10, and the spatial intelligence gathering system 10 may receive the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100 prior to monitoring the environment for electromagnetic EM signals of interest.

[0101] As in the embodiments discussed above with reference to Figures la and lb, the one or more robots 200a to 200n may be released into the environment 100 (or continue to be present in the environment 100 after the generation of the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100) and move around the environment in a closed loop manner. Here, each one of the one or more robots 200a to 200n may direct their own and each other's movements around the environment 100 by inter-robot communication (using, for example, the at least one antenna 220 on each robot), based on EM signals detected in the environment 100 by the EM sensor 230h (and/or the other sensors of the sensors 230) of the spatial intelligence node 50 on each robot. Similarly, such directed movement of the one or more robots 200a to 200n may be based on the absence of an EM signal(s) in a given area of the environment 100, where the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 indicate that such an EM signal(s) should be present.

[0102] As the one or more robots 200a to 200n move around the environment 100, the spatial intelligence gathering system 10 assesses whether any detected EM signals can be characterised (or identified) as "signals of interest". In some embodiments, a signal of interest 70 may be referred to as an "anomalous signal" or a "signal requiring investigation".

[0103] Here, a signal of interest 70 may be an EM signal that is not expected to be present in the environment 100. For example, in some embodiments a signal of interest 70 may be a detected EM signal that is stronger or weaker than expected for the environment 100, at a frequency or frequencies that would not be expected for the environment 100, is intermittent in a manner that is not expected for the environment 100, has not been detected previously (e.g. during earlier periods of monitoring of the environment 100 by the spatial intelligence gathering system 10), has been detected previously but now has different EM characteristics (e.g. frequency, signal strength, etc.), has an unexpected or unknown modulation or coding scheme, or a combination of any of the above. In some embodiments, a signal of interest 70 may be an EM signal that has been previously detected (e.g. during a previous monitoring session, or during the generation of the map of the physical layout 110 of the environment 100 and the electromagnetic model 120) but is now absent from the environment 100. In such scenarios, the spatial intelligence gathering system 10 may characterise or identify the absence of the previously detected EM signal as a "signal of interest".

[0104] In some embodiments, a detected EM signal may be characterised as a signal of interest 70 as a consequence of the continued detection of that EM signal over a period of time. For example, a detected EM signal may not be characterised as a signal of interest 70 if detected over short time period (e.g. over a few minutes or over an hour), but may later be characterised as a signal interest 70 if it is still detected over a longer time period (e.g. many hours or days). Similarly, a detected EM signal may be ii characterised as a signal of interest 70 based on the time of day that it is detected. For example, such a detected EM signal may not be characterised as a signal of interest 70 if it detected during the day (e.g. between 9am and 5pm), but may be characterised as a signal of interest 70 if it is detected at night (e.g. between lam and 6am).

[0105] It will be appreciated that various criteria may be employed to categorise a detected EM signal as being a signal of interest 70, based on the specific circumstances of the scenario in question (e.g. the nature of the particular environment 100 in question, and/or the specific electronic devices and electronic components present, or expected to be present, in the particular environment 100 in question). It will also be understood that any of the above discussed criteria for characterising a detected EM signal as a signal of interest 70 may be used in combination with any other criteria discussed above, or any other suitable criteria.

[0106] A detected EM signal may be categorised as being a signal of interest 70 using one or both of the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100, as well as or as an alternative to the above discussed considerations. For example, a detected EM signal may be detected in an area of the environment 100 in which no EM signal is expected to be present, where no electronic devices or electrical components are known to be present based on the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100. As a further example, a detected EM signal may be at a frequency or frequencies that do not correspond to any electrical device or electrical components (e.g. a desktop computer, wireless hub, or electrical socket) known to be present in an area of the environment 100 based on the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100.

[0107] Similarly, an EM signal may be detected in an area of the environment 100 where the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100 indicate that an EM signal emitted by an electrical device or electrical component would be expected to be detected. That is to say, it may be determined that a detected EM signal is likely to be associated with a known electrical device or electrical component located in the environment 100, based on the characteristics of the detected EM signal and the expected EM behaviour in that area of the environment 100 as indicated by the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100.

[0108] For example, an EM signal may be detected in an area of the environment 100 and determined as being likely to be associated with a desktop computer known to be located in the environment 100. This determination may be based on the characteristics of the detected EM signal (e.g. signal strength, frequency, signal repetition rate, etc.)

/ 1 compared with the expected EM behaviour of EM signals emitted by that desktop computer in that area of the environment 100 (e.g. how an EM signal emitted by the desktop computer would be expected to propagate through the environment 100 to the area at which the detected EM signal was detected). It will be appreciated that the detected EM signal may not necessarily be detected in the same area of the environment 100 at which the desktop computer is known to be located.

[0109] In such scenarios, the spatial intelligence gathering system 10 may conclude that the detected EM signal is not a signal of interest 70, based on that detected EM signal having been determined as being likely to correspond to a known electrical device or electrical component located in the environment 100 (using the characteristics of the detected EM signal and the expected EM behaviour in that area of the environment 100 as indicated by the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100).

[0110] In some embodiments, the spatial intelligence gathering system 10 may determine that a detected electromagnetic signal is unusual for the environment 100, but nonetheless should not be characterised as a signal of interest 70. For example, the spatial intelligence gathering system 10 may detect an electromagnetic signal that has not been detected during previous monitoring sessions, but nonetheless should not be characterised as a signal of interest 70 since that detected electromagnetic signal is consistent with an expected eventuality for the environment 100 (e.g. an individual being present in the environment with a mobile phone, where the detected electromagnetic signal is consistent with a mobile phone signal).

[0111] A detected EM signal may be characterised (or categorised, or identified) as a signal of interest 70 by one or more robots of the one or more robots 200a to 200n_z and/or by the server 250 (in embodiments where the server 250 is included), or by any other aspect of the spatial intelligence gathering system 10.

[0112] In some embodiments, the spatial intelligence gathering system 10 may utilise one or more machine learning algorithms to characterise a detected electromagnetic signal as a signal of interest 70, where such machine learning algorithms have previously been trained to characterise signals as such.

[0113] In some embodiments, the spatial intelligence gathering system 10 may categorise detected signals as signals of interest 70 in a dynamic manner, as the one or more robots 200a to 200n continue to move around the environment 100. In some embodiments, the spatial intelligence gathering system 10 may categorise detected signals as signals of interest after the one or more robots 200a to 200n have finished moving around the environment 100 (i.e. by storing the detected signals, along with the locations where those signals where detected, for processing later). [0114] Returning to Figure 2a, robots 200a, 200b, and 200c, are released into the environment 100 to monitor for and detect electromagnetic (EM) signals. In the example scenario of Figure 2a, EM signals 60 are detected by one or more of the robots 200a, 200b, and 200c from several electrical devices and electrical components in the environment 100, in particular from a computer desktop screen 105e, a telephone 105f, and from multiple light switches 105h (shown in Figure 2a as curved lines emitting from each of these electrical devices and electrical components). As in the case of Figure lb above, reflections, refractions, and scatters off of, or transmissions through, the various surfaces in the environment 100 are shown in Figure 2a by dotted curved lines. However, the characteristics of the detected EM signal 60 emitted from location A may cause one or more of the robots 200a, 200b, and 200c of the spatial intelligence gathering system 10 to characterise that EM signal 60 as a signal of interest 70. It is highlighted that the one or more of robots 200a, 200b, and 200c may detect and categorise the EM signal as a signal of interest 70 without being able to ascertain that the detected signal of interest 70 was emitted from location A (e.g. if the signal of interest 70 was reflected, refracted, scattered off of, and/or transmitted through, the various surfaces in the environment 100 before being detected).

[0115] In some embodiments, once a detected signal has been characterised as (or identified as, determined to be, or categorised as) a signal of interest 70, the spatial intelligence gathering system 10 may output the details of the signal of interest 70 to an external system or user for further evaluation, investigation, repair, or other action. [0116] After a signal of interest 70 has been identified in the environment 100, in some embodiments the one or more robot of the one or more robots 200a to 200n may be directed around the environment 100 based on the identified signal of interest 70. For example, the identified signal of interest 70 may be stored, along with the location in the environment 100 where the source of the signal of interest 70 was identified as originating from, and the one or more robots 200a to 200n may then continue to move around the environment 100 to identify further signals of interest. The location(s) of the one or more robots 200a to 200n in the environment 100 when the source of the signal of interest 70 was identified by the one or more robots 200a to 200n, or when the EM data used in the identification was collected by those one or more robots 200a to 200n_z may also be stored.

[0117] In some embodiments, the one or more robots 200a to 200n may be directed around the environment 100 in order to identify the source of the signal of interest 70 in the environment 100.

[0118] To continue the example scenario of the Figures above, Figure 2b shows three robots 200a to 200c being directed around the environment 100 based on the identified signal of interest 70, and may be directed around the environment 100 in order to identify the source of that signal of interest 70.

[0119] As in the embodiments discussed above with reference to Figures la, lb, and 2a, the one or more robots 200a to 200n may move around the environment in a closed loop manner. Here, each one of the one or more robots 200a to 200n may direct their own and each other's movements around the environment 100 by inter-robot communication (using, for example, the at least one antenna 220 on each robot), based on the source of a signal of interest 70 in the environment 100.

[0120] In some embodiments, the one or more robots 200a to 200n may move around the environment 100 based on the signal of interest 70, for example to improve detection of the signal of interest 70 and/or to gather further information regarding the signal of interest 70 (e.g. to record certain characteristics of the signal with greater accuracy).

[0121] In some embodiments where a plurality of robots 200a to 200n are included in the spatial intelligence gathering system 10, the plurality of robots 200a to 200n may communicate with each other (and, if included, the server 250) to assign one or more robots of the plurality of robots 200a to 200n to identify or determine the source of a detected signal of interest 70. For example, the plurality of robots 200a to 200n may self-organise to assign the one or more robots to identify or determine the source of a detected signal of interest 70, or this may be done by a "lead" robot, or may be done by the server 250 (where included). This may be achieved, for example, using the communications module 240 of the spatial intelligence node 50 of each robot.

[0122] In some embodiments where a single robot of the one or more robots 200a to 200n is assigned to move based on the signal of interest 70, or to identify or determine the source of the signal of interest 70, the one robot of the one or more robots 200a to 200n may move around the environment 100 based upon a continued detection of the signal of interest 70. For example, the one of the one or more robots 200a to 200n may move through the environment 100 in a manner where the detected signal of interest 70 increases in strength (i.e. moving up the gradient of detected signal intensity), until the one of the one or more robots 200a to 200n reaches a location in the environment 100 where the signal strength of the detected signal of interest 70 is at a maximum.

[0123] In some embodiments, the one or more robots 200a to 200n that are assigned to identify or determine the source of a signal of interest 70 may move around the environment 100 to identify or determine the source of the signal of interest 70 based on the electromagnetic model 120 of the environment 100. Here, the electromagnetic model 120 of the environment 100 may be used to predict how the environment 100 may cause the signal of interest 70 to propagate through the environment 100 (e.g. by reflecting, diffracting, transmitting, or scattering the signal of interest 70), or how the signal of interest 70 may be interacting with other electromagnetic signals in the environment 100.

[0124] For example, the characteristics of the signal of interest 70 (e.g. signal strength, phase, angle of arrival spectrum, polarisation, or other characteristic), as detected by each of the one or more robots 200a to 200n_z may be used with the electromagnetic model 120 to produce a probability field for the location of the source of the signal of interest 70 in the environment 100. Such a probability field may then be updated as the one or more robots 200a to 200n move through the environment 100 to identify the source of the signal of interest 70, until a sufficient level of accuracy in the location of the identified source is reached. It will be noted that the one or more robots 200a to 200n may generate a probability field individually, based on the characteristics of the signal of interest 70 detected by that robot in combination with the electromagnetic model 120, which can then be combined with the probability field of other robots of the one or more robots 200a to 200n (where multiple robots are assigned to identify the source of a signal of interest 70). Alternatively, the spatial intelligence gathering system 10 may generate a single probability field, based on the characteristics of the signal of interest 70 as detected by each of the robots in combination with the electromagnetic model 120.

[0125] As such, the electromagnetic model 120 may be used by the spatial intelligence gathering system 10 to account for reflections, diffractions, or scatters, as well as any other electromagnetic interactions, with the environment 100 in order to predict the source of the signal of interest 70, and direct the one or more robots 200a to 200n to identify or determine the source (or a possible source) of the signal of interest 70.

[0126] In some embodiments where a plurality of the one or more robots 200a to 200n are assigned to identify the source of the signal of interest 70, the plurality of robots may move through the environment 100 by directing each other based on the signal of interest 70, or to identify the source of the signal of interest 70 in the environment 100. For example, the plurality of the one or more robots 200a to 200n may communicate with each other (e.g. using the communications modules 240 and the at least one antenna 220 of each robot) to direct each other's movement in order to identify a location at which the signal strength of the signal of interest 70 is at a maximum (e.g. by employing interpolation or trilateration/multilateration methods).

[0127] It will be understood that any characteristic of the detected signal of interest 70 may be used to direct the one or more robots 200a to 200n to identify or determine the source of the detected signal of interest 70, and signal strength is used only as a nonlimiting example.

[0128] In some embodiments, the plurality of robots may employ angle of arrival, AoA, and/or time difference of arrival, TDoA, techniques to identify the source of the signal of interest 70. Since each of the plurality of robots knows its location (e.g. x, y, z coordinates) within the environment 100 (and, in some scenarios, the orientation of the at least one antenna 220 and/or one or more sensors 230, which may be variable using an actuator or motor as discussed above), AoA and/or TDoA techniques may be employed between the plurality of robots to identify the source of the signal of interest 70 in the environment 100 (i.e. through inter-robot communication as the plurality of robots move through the environment 100). For example, the ratios of the signal strengths as received by each of the plurality of robots may be used to identify the source of the signal of interest 70 in the environment 100 relative to each of the plurality of robots.

[0129] As a further example, in those embodiments where the spatial intelligence node 50 of each of the plurality of robots includes multiple antennas, AoA and/or TDoA techniques may be employed by each robot using the multiple antennas of its spatial intelligence node 50, such that each robot may individually identify the source of the signal of interest 70. This may be used separately (e.g. in embodiments where a single robot is used to identify the source of a signal of interest 70), or in addition to AoA and/or TDoA techniques employed between the plurality of robots (i.e. by inter-robot communication).

[0130] In some embodiments, the one or more robots 200a to 200n may identify the source of the signal of interest 70 by utilising the electromagnetic model 120 of the environment 100, for instance by using the electromagnetic model 120 to predict the location of the signal of interest 70 based on the characteristics of the signal of interest 70 at the location where it was first detected. This approach may be used in combination with other approaches discussed herein, and may be employed in an iterative manner. For example, the electromagnetic model 120 may be used repeatedly to model potential sources for the signal of interest 70, as the signal of interest 70 is continually or periodically detected as the one or more robots move to identify the source of the signal of interest 70 and/or cycle through different sensing frequencies.

[0131] It will be understood that the source of a signal of interest 70 in the environment 100 might be identified or determined through a variety of different approaches, where the one or more robots 200a to 200n may be utilised or not, and the above examples are not intended to be limiting. [0132] In some embodiments, the one or more robots 200a to 200n directed to identify or determine the source of a signal of interest 70 in the environment 100 may utilise one or both of the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100. This may be in addition to, or as an alternative to, any approach discussed above.

[0133] For example, one or both of the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100 may be used to predict the source of the detected signal of interest 70, by using the expected EM behaviours for the environment 100 indicated by the electromagnetic model 120 of the environment 100.

[0134] In some embodiments, the spatial intelligence gathering system 10 may identify priority locations in the environment 100 that, if a signal of interest 70 is detected near such a priority location, that priority location is prioritised as a possible source of the signal of interest 70. For example, such an identified priority location may be a cupboard in the environment 100, and if a signal of interest 70 is detected within a threshold distance from that priority location (i.e. the cupboard), then the one or more robots assigned to identify the source of the signal of interest 70 may prioritise investigating that priority location (i.e. the cupboard) to ascertain if it is the source of the signal of interest 70. Such priority locations in the environment 100 may be identified using one or both of the map of the physical layout 110 and the electromagnetic model 120 of the environment 100, or by using a machine learning algorithm trained to identify possible priority locations (e.g. using one or both of the map of the physical layout 110 and the electromagnetic model 120 of the environment 100, and/or using sensor data recorded by the spatial intelligence node 50 of each robot). Alternatively or additionally, such priority locations may be manually selected by a user.

[0135] It will be understood that, in each embodiment discussed above, the identification or determination of the source of a signal of interest 70 may be an estimation of the location of the source of the signal of interest 70 in the environment 100. For example, the identified or determined source of the signal of interest 70 may correspond to an area or volume of the environment 100, rather than a precise location. As such, the spatial intelligence gathering system 10 may be configured to identify or determine the source of a signal of interest 70 only to a certain level or threshold of accuracy or certainty (e.g. identifying the source to within 1 metre, 1 cm, or any other suitable distance).

[0136] In some embodiments, once the source in the environment 100 of a detected signal of interest 70 has been identified or determined, the spatial intelligence gathering system 10 may determine if a known electrical device or component corresponds to that identified source of the detected signal of interest 70. For example, the spatial intelligence gathering system 10 may determine that the source of the detected signal of interest 70 is an electrical socket located in the environment 100, and that the detected signal of interest 70 is a signal with characteristics that would not be expected to be emitted from an electrical socket (e.g. an unusual signal frequency, or signal repetition frequency).

[0137] In some embodiments where a plurality of robots 200a to 200n are utilised, only some of the plurality of robots 200a to 200n may be assigned to identify or determine the source of a signal of interest 70 in the environment 100. At the same time, other robots of the plurality of robots 200a to 200n may continue to move around the environment 100 and detect other electromagnetic signals, in order to assess whether such signals should be characterised as a signal of interest 70.

[0138] Returning to Figure 2b, robots 200a, 200b, and 200c move around the environment 100 to identify the source of the signal of interest 70. Here, one or more of the robots 200a, 200b, and 200c may detect the signal of interest 70 as it propagates through the environment 100 (e.g. through reflections, refractions, scatters off of, or transmissions through, the various surfaces in the environment 100), and use those detected signals, with the electromagnetic model 120 of the environment 100, to ascertain the likely source of the signal of interest 70. It is highlighted that any of the other methods discussed above for identifying the source of the signal of interest 70 might be applied to the example scenario of Figure 2b. In the example scenario of Figure 2b, each of robots 200a, 200b, and 200c identify the source of the signal of interest 70 as being location A but, as discussed above, this may be carried out by only some (or only a single) robot of the spatial intelligence gathering system 10.

[0139] In some embodiments, once the source in the environment 100 of a detected signal of interest 70 has been identified or determined, the spatial intelligence gathering system 10 may output the details of the detected signal of interest 70, and its source in the environment 100, to an external system or user for further evaluation, investigation, repair, or other action. For example, the source in the environment 100 of the detected signal of interest 70 may indicate that an electrical device in the environment is faulty, that a newly introduced electrical device is interfering with other electrical devices in the environment, or that an unauthorised electrical device has been introduced into the environment.

[0140] In some embodiments, once the source in the environment 100 of a detected signal of interest 70 has been identified or determined, the one or more robots of the one or more robots 200a to 200n that were assigned to identify or determine that source may return to monitoring the environment 100. Here, the one or more robots may return to detecting EM signals in the environment 100 (e.g. may return to a designated patrol route around the environment 100), and assessing whether those detected EM signals might be characterised or identified as being signals of interest 70 (or characterised as not being signals of interest, based on the assessment discussed above). For example, the one or more robots 200a to 200n that were assigned to identify or determine the source of the signal of interest 70 may then be directed based upon a further detected signal (which may be characterised as a further signal of interest 70).

[0141] Fig ure 3 shows a flowchart of a method 1000 for the spatial intelligence gathering system 10 to gather intelligence on an environment 100 according to some embodiments of the present application.

[0142] It will be understood that any of the processes and functionalities discussed above may be applied in conjunction with, or as equivalents of, the steps of the flowchart shown in Figure 3.

[0143] Some of the steps shown in Figure 3 are optional steps (that is to say, not essential to work the invention of the present application). Such optional steps are indicated as such with a dotted line.

[0144] It will also be understood that Steps 1100 and 1200, concerning the generation of a map of the physical layout 110 of the environment 100 and the generation or population of the electromagnetic model 120 of the environment 100, may be carried out separately to the other steps, or may be omitted entirely (e.g. when a system external to the spatial intelligence gathering system 10 carries out those steps or their equivalent). This is denoted in the flowchart of Figure 3 by dotted line 1050.

[0145] At Step 1100, one or more robots 200a to 200n are released into an environment 100. The one or more robots 200a to 200n move around the environment 100 in a closed loop manner, where movement may be directed through inter-communication between multiple robots 200a to 200n_z a single robot directing its own movement, and/or between the one or more robots 200a to 200n and a server 250 (if included) of the spatial intelligence gathering system 10 (e.g. based on spatial sensor data recorded by the spatial intelligence node 50 of each of the one or more robots 200a to 200n).

[0146] The one or more robots 200a to 200n may move around the environment 100 in order to generate or populate a map of the physical layout 110 of the environment 100. The map of the physical layout 110 of the environment 100 may be generated or populated using a spatial intelligence node 50 included on each of the one or more robots 200a to 200n_z the spatial intelligence node 50 including one or more sensors 230 used for the generating or populating. The one or more sensors 230 may include sensors such as a camera 230a, a 2D and/or 3D lidar sensor 230b, an infrared sensor 230c, a thermal sensor 230d, a depth camera 230e, a hyperspectral camera 230f, and/or a sonar sensor 230g.

[0147] The one or more robots 200a to 200n may dynamically update the map of the physical layout 110 of the environment 100 as the one or more robots 200a to 200n move around the environment 100. The one or more robots 200a to 200n may generate or populate the map of the physical layout 110 of the environment 100 by utilising a simultaneous localization and mapping (SLAM) approach. Alternatively, or additionally, other suitable approaches may be used, for example by using a global navigation satellite system, GNSS, or by using fiducial markers, ultrasound beacons, Bluetooth low energy beacons, and so on.

[0148] The one or more robots 200a to 200n may share the map of the physical layout 110 of the environment 100 with other robots of the one or more robots 200a to 200n as it is updated, and the movement of the one or more robots 200a to 200n may be directed based on the shared and dynamically updated map of the physical layout 110 of the environment 100.

[0149] At Step 1200, the one or more robots 200a to 200n may move around the environment 100 to generate or populate an electromagnetic model 120 of the environment 100. Here, the movement of the one or more robots 200a to 200n may be in a closed loop manner, where movement may be directed through intercommunication between multiple robots 200a to 200n_z by a single robot directing its own movement, and/or between the one or more robots 200a to 200n and a server 250 (where included) of the spatial intelligence gathering system 10.

[0150] The electromagnetic model 120 of the environment 100 may be generated or populated using one or more of the spatial intelligence nodes 50 included on each of the respective one or more robots 200a to 200n_z where the spatial intelligence node 50 includes one or more EM sensors 230h (as well as the other sensors of the sensors 230) used for the generating or populating.

[0151] The one or more robots 200a to 200n (and/or, where included, the server 250) may dynamically update the electromagnetic model 120 of the environment 100 as the one or more robots 200a to 200n move around the environment 100. The one or more robots 200a to 200n (and/or, where included, the server 250) may share the electromagnetic model 120 of the environment 100 with other robots of the one or more robots 200a to 200n as it is updated, and the movement of the one or more robots 200a to 200n may be directed based on the shared and dynamically updated electromagnetic model 120 of the environment 100. The one or more robots 200a to 200n (and/or, where included, the server 250) may generate or populate the electromagnetic model 120 of the environment 100 by utilising a SLAM approach in an equivalent manner to discussed above with respect to Step 1100.

[0152] The electromagnetic model 120 of the environment 100 may take the form of a channel model, the channel model being a representation of the effects of a communication channel (e.g. a medium or environment) through which wireless signals are propagated (as discussed above).

[0153] At Step 1300, the one or more robots 200a to 200n may be re-released into the environment 100 or, if already in the environment 100, continue to move around the environment 100 in order to monitor and detect EM signals present in the environment 100.

[0154] Here, the movement of the one or more robots 200a to 200n may be in a closed loop manner, where movement is directed through inter-communication between the one or more robots 200a to 200n_z and/or between the one or more robots 200a to 200n and a server 250 of the spatial intelligence gathering system 10. Such directed movement may be for the purpose of increasing the efficiency of the monitoring and detecting of EM signals present in the environment 100 (for example with respect to power consumption, detection rate of signals of interest 70, and timely investigation of signals of interest 70).

[0155] At Step 1400, the spatial intelligence gathering system 10 may characterise or identify an EM signal detected by the one or more robots 200a to 200n as being a signal of interest 70. The EM signal detected by the one or more robots 200a to 200n may be characterised or identified as being a signal of interest 70 by the one or more robots 200a to 200n that detected the EM signal (i.e. the characterisation may be made by the one or more robots themselves), or may be characterised or identified by other parts of the spatial intelligence gathering system 10 (e.g. the server 250, where included).

[0156] The detected EM signal may be characterised or identified as a signal of interest 70 based on the characteristics of that detected EM signal. For example, the detected EM signal may be stronger or weaker than expected for the environment 100, may be at a frequency or frequencies that would not be expected for the environment 100, may be intermittent in a manner that is not expected for the environment 100, may have not been detected previously (e.g. during earlier periods of monitoring of the environment 100 by the spatial intelligence gathering system 10), or may be a combination of any of the above or any other characteristics of the detected EM signal (including the characteristics discussed above).

[0157] In addition to, or as an alternative to, the above discussed considerations, the detected EM signal may be categorised or identified as being a signal of interest 70 using one or both of the map of the physical layout 110 of the environment 100 and

Bi the electromagnetic model 120 of the environment 100. For example, where the detected EM signal is detected in an area where no EM signals with those signal characteristics would be expected to be detected, based on the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100.

[0158] At optional Step 1450, details of the characterised or identified signal of interest 70 may be output to an external system or user for further evaluation, investigation, repair, or other action.

[0159] At Step 1500, the one or more robots 200a to 200n may move around the environment 100 based on the characterised or identified signal of interest 70. For example, one or more of the one or more robots 200a to 200n may move to improve detection of the signal of interest 70 and/or to gather further information regarding the signal of interest 70 (e.g. to record certain characteristics of the signal with greater accuracy).

[0160] The movement of the one or more robots 200a to 200n may be directed in order for the spatial intelligence gathering system 10 to identify or determine the location of the source in the environment 100 of the characterised or identified signal of interest 70. Here, the movement of the one or more robots 200a to 200n may be in a closed loop manner, where movement is directed through inter-communication between the one or more robots 200a to 200n_z and/or between the one or more robots 200a to 200n and a server 250 of the spatial intelligence gathering system 10. Such directed movement may be for the purpose of increasing the efficiency of identifying the location or source of the signal of interest 70 present in the environment 100 (for example with respect to power consumption, detection rate of the signal of interest 70, and timely investigation of signal of interest 70).

[0161] Where a plurality of robots 200a to 200n are included in the spatial intelligence gathering system 10, the plurality of robots 200a to 200n may communicate with each other (and, if included, the server 250) to assign one or more robots of the plurality of robots to identify or determine the source of the detected signal of interest 70. This may be achieved, for example, using the communications module 240 of the spatial intelligence node 50 of each robot.

[0162] Where a single robot of the one or more robots 200a to 200n is assigned to identify or determine the location of the signal of interest 70, the one robot of the one or more robots 200a to 200n may move to identify the source of the signal of interest 70 in the environment 100 based upon a continued detection of the signal of interest 70. [0163] The one of the one or more robots 200a to 200n assigned to identify or determine the location of the signal of interest 70 may move through the environment 100 such that the detected signal of interest 70 increases in strength (i.e. moving up the gradient of detected signal intensity), until a location is reached in the environment 100 where the signal strength of the signal of interest 70 is at a maximum.

[0164] Where a plurality of the one or more robots 200a to 200n are assigned to identify the source of the signal of interest 70, the plurality of robots may move through the environment 100 by directing each other to identify the source of the signal of interest 70 in the environment 100. Here, the plurality of the one or more robots 200a to 200n may communicate with each other to direct each other's movement in order to identify a location at which the signal strength of the signal of interest 70 is at a maximum (e.g. by employing interpolation, trilateration, or other multilateration methods). This may be achieved, for example, using the communications module 240 of the spatial intelligence node 50 of each robot.

[0165] Alternatively, or additionally, the one or more robots 200a to 200n assigned to identify the source of the signal of interest 70 may employ angle of arrival, AoA, and/or time difference of arrival, TDoA, techniques to identify the source of the signal of interest 70 in the manner discussed above.

[0166] The one or more robots 200a to 200n directed to identify or determine the source of a signal of interest 70 in the environment 100 may utilise one or both of the map of the physical layout 110 of the environment 100 and the electromagnetic model 120 of the environment 100. This may be in addition to, or as an alternative to, any approach discussed above.

[0167] At optional Step 1550, details of the signal of interest 70 and/or the location of the signal of interest 70 may be output to an external system or user for further evaluation, investigation, repair, or other action.

[0168] Optionally, after Step 1500 (or, where relevant, after Step 1550), the method 1000 may at Step 1600 return to Step 1300 and the one or more robots 200a to 200n may continue to detect EM signals in the environment 100. In the flowchart of Figure 3, this is denoted by a dotted line.

[0169] The above discussion and method 1000 allows for the spatial intelligence gathering system 10 to map an environment 100 and model the electromagnetic behaviour of that environment 100 to a suitable level of accuracy. This can then be used to identify unusual or unexpected signals in that environment 100, or track changes in the electromagnetic behaviour of the environment 100 over time. For example, the one or more robots may be released periodically into the environment 100, where a suitably accurate map 110 and electromagnetic model 120 of the environment 100 has previously been generated (e.g. by the spatial intelligence gathering system 10 as discussed above, or otherwise), to monitor for new signals of interest or to track changes in the electromagnetic behaviour of the environment 100 over time.

[0170] It will be highlighted that, where a plurality of robots are employed, not all of the robots may be released or deployed into the environment 100 at the same time. For example, a single robot may monitor or patrol the environment 100 to characterise detected EM signals as signals of interest 70, and the other robots of the plurality of robots may only be deployed in response to a detected EM signal being characterised as a signal of interest 70. Here, the other robots of the plurality of robots may otherwise be on standby (e.g. at a docking station(s)) or engaged with other tasks separate to the spatial intelligence gathering system 10 (e.g. as cleaning robots or otherwise patrolling the environment 100).

[0171] In the context of ongoing periodic patrolling by the one or more robots 200a to 200n_z signals of interest 70 may be detected over a short-term search (e.g. less than a day) or through changes in detection against a longer-term characterisation of the electromagnetic behaviour of a building (or the radio frequency "pattern of life" of that building) (e.g. over a period of a week or more).

[0172] This allows for the spatial intelligence gathering system 10 to be applied to a variety of different purposes, such as for security applications (e.g. counter eavesdropping), the management of complex wireless networks (such as radio coverage surveys (mobile, satellite, television, etc.)), or electronic device monitoring. Similarly, the spatial intelligence gathering system 10 may be used for obtaining a "fingerprint" of a physical structure, such as the characteristic interaction of that structure with the local radio frequency or electromagnetic environment, which commonly will comprise various active electronic sources (e.g. Wi-Fi, loT, cellular, etc.).

[0173] Other scenarios in which the above approaches might be applied include archaeological or geological surveys, where physical and electromagnetic mapping of a given site would be faster, more efficient, and more accurate than might be otherwise achieved through manual or static methods.

[0174] The above discussion regarding the generation or population of the electromagnetic model 120 of the environment 100 concerns "passive" mapping, where the one or more robots 200a to 200n record EM data relating to EM signals already present in the environment 100, using the one or more EM sensors 230h (and/or the other sensors of the sensors 230) of the spatial intelligence node 50 on at least one of the one or more robots 200a to 200n. [0175] Alternatively, or in addition to this approach discussed above, in some embodiments the one or more robots 200a to 200n may map the electromagnetic behaviour of the environment 100 by actively emitting and detecting EM signals.

[0176] Fig ure 4 shows a scenario according to those embodiments, where the plurality of robots 200a to 200n generate or populate the electromagnetic model 120 of the environment 100 by "active" mapping of the electromagnetic behaviour of the environment 100.

[0177] As in the example scenarios of the Figures above, Figure 4 shows three robots 200a, 200b, and 200c being directed around the environment 100 to generate or populate the electromagnetic model 120 of the environment 100 by active mapping of the electromagnetic behaviour of the environment 100.

[0178] As in the embodiments discussed above with reference to Figures la, lb, 2a, and 2b, the plurality of robots 200a to 200n may move around the environment 100 in a closed loop manner. Here, each one of the plurality of robots 200a to 200n may direct their own and each other's movements around the environment 100 by inter-robot communication (using, for example, the at least one antenna 220 on each robot), in order to generate or populate the electromagnetic model 120 of the environment 100. [0179] In a similar manner to the embodiments discussed above, in the present embodiments the plurality of robots 200a to 200n may move around the environment 100 and generate or populate the electromagnetic model 120 of the environment 100 by utilising simultaneous localization and mapping (SLAM) techniques. [0180] One or more robots of the plurality of robots 200a to 200n of the present embodiments may move around the environment 100 and generate or populate the electromagnetic model 120 of the environment 100 by periodically emitting an electromagnetic signal 300, where each emitted electromagnetic signal 300 is emitted at a respective emission location in the environment 100. Here, each emitted electromagnetic signal 300 may be emitted using the at least one antenna 220, may be emitted using the one or more sensors 230, or may be emitted using a different component of the one or more robots in question.

[0181] Here, each emitted electromagnetic signal 300 comprises a set of emission characteristics (e.g. signal strength, frequency profile, duration, pulse rate, or any other suitable parameter of an electromagnetic signal).

[0182] In some embodiments, the emitted electromagnetic signal 300 may include an identification parameter (e.g. an identification number, code, or other ID) for the robot that emitted the electromagnetic signal 300, as well as data indicating the location in the environment 100 at which the electromagnetic signal 300 was emitted by that robot (i.e. the emission location). The emitted electromagnetic signal 300 may additionally include data indicating the time at which the electromagnetic signal 300 was emitted (i.e. the emission time).

[0183] In some embodiments, any of the identification parameter, emission location, and emission time may be part of the electromagnetic signal 300 being emitted, or may be transmitted or communicated to the other robots of the plurality of robots 200a to 200n separately to the electromagnetic signal 300. In some embodiments, any of the identification parameter, emission location, and emission time may be transmitted or communicated to the other robots of the plurality of robots 200a to 200n at the same time that the electromagnetic signal 300 is emitted, before the electromagnetic signal 300 is emitted, or after the electromagnetic signal 300 is emitted. This communication may be achieved, for example, using the communications module 240 of the spatial intelligence node 50 of each robot.

[0184] One or more of the other robots of the plurality of robots 200a to 200n may then detect (or receive) the electromagnetic signal 300 emitted by the one or more robots. Here, the one or more of the other robots that detects (or receives) the electromagnetic signal 300 is located at a different location in the environment 100 to that of the robot that emitted the electromagnetic signal 300 (i.e. the emission location). The robot that emitted the electromagnetic signal 300 may itself also detect (or receive) the electromagnetic signal 300 (i.e. through reflection, refraction, or scattering in the environment 100).

[0185] The electromagnetic signal 300 that is detected (or received) by each of the one or more of the other robots of the plurality of robots 200a to 200n may have a set of received characteristics (e.g. signal strength, frequency profile, duration, or any other suitable parameter of an electromagnetic signal) for each robot that detects (or receives) the electromagnetic signal 300.

[0186] The set of received characteristics of the electromagnetic signal 300 that is detected (or received) by each of the one or more of the other robots of the plurality of robots 200a to 200n may be different to the set of emission characteristics of the emitted electromagnetic signal 300 as a consequence of the emitted electromagnetic signal 300 having propagated through the environment 100 from the emission location. For example, the set of received characteristics may have a different signal strength, frequency profile, duration, or any other suitable parameter compared with the set of emission characteristics.

[0187] In some embodiments, each of the one or more of the other robots of the plurality of robots 200a to 200n that detected (or received) the electromagnetic signal 300 may then communicate or share the set of received characteristics, as well as the location in the environment 100 at which the electromagnetic signal 300 was detected (or received), with the other robots of the plurality of robots 200a to 200n (and/or, where included, the server 250). This communication may be achieved, for example, using the communications module 240 of the spatial intelligence node 50 of each robot.

[0188] The spatial intelligence gathering system 10 may then generate or populate the electromagnetic model 120 of the environment 100 based on the set of emission characteristics compared with the sets of received characteristics for the electromagnetic signal 300.

[0189] For example, the received characteristics of the electromagnetic signal, when compared with the emission characteristics, may indicate that a signal strength of the electromagnetic signal 300 has reduced by a given amount as the electromagnetic signal 300 has propagated through the environment 100. This reduction may then be used, in combination with the time and location of emission of the electromagnetic signal 300 and the time and location of receipt of the electromagnetic signal 300, to model how the electromagnetic signal 300 propagated through the environment 100.

[0190] In some embodiments, the emission characteristics of the electromagnetic signal 300 may be adjusted between emissions based on the physical layout 110 or the EM model 120 of the environment 100 (e.g. as the electromagnetic model 120 of the environment 100 is generated or populated). For example, a pulse rate of the electromagnetic signal 300 may be adjusted to account for the physical materials of objects and structures in the environment 100, and/or whether the robot emitting the electromagnetic signal 300 has placed the emitter in contact with the surface (e.g. in embodiments where the emitting antenna, which may be the at least one antenna 220 or EM sensor 230h, is installed on an actuator arm).

[0191] It will be appreciated that, in some embodiments, the above approach may be employed using a single robot, where that robot emits the electromagnetic signal 300 and detects or receives the electromagnetic signal 300 after it has been scattered by (or otherwise propagated through) the environment 100. In such embodiments, the electromagnetic model 120 of the environment 100 may be generated or populated based on the set of emission characteristics for the electromagnetic signal 300 as emitted by the single robot, compared with the set(s) of received characteristics for the electromagnetic signal 300 as received by the single robot.

[0192] Furthermore, it will be appreciated that, in some embodiments the electromagnetic signal 300 may be emitted by a different aspect of the spatial intelligence gathering system 10, such as from a device in a static position within the environment 100 selected by a user, by a device held by a user and moved around the environment 100, or from on a non-autonomous mobile platform (e.g. a moving car, boat, plane, or other vehicle, manned or unmanned).

[0193] In this manner, the electromagnetic model 120 of the environment 100 may be generated or populated through repeated emission and receipt of electromagnetic signals 300 at different locations in the environment 100. This may be in combination with any of the methods discussed above, and may be additionally based on the generated map of the physical layout 110 of the environment 100.

[0194] In some embodiments, each of the robots of the plurality of robots 200a to 200n_z a subset of the plurality of robots 200a to 200n_z and/or the server 250 (where included) may generate or populate the electromagnetic model 120 of the environment 100 in this manner based on the set of emission characteristics compared with the sets of received characteristics for the electromagnetic signal.

[0195] In some embodiments, only a single robot of the plurality of robots 200a to 200n may emit electromagnetic signals 300 at various locations in the environment 100, with the other robots of the plurality of robots 200a to 200n "listening" for the electromagnetic signals 300. In those embodiments, the electromagnetic signals 300 may have predetermined emission characteristics that can be recognised by each of the "listening" robots, allowing identification of the detected (or received) signal as being the electromagnetic signal 300.

[0196] In some embodiments, each of the plurality of robots 200a to 200n may periodically emit an electromagnetic signal 300 to be detected by others of the plurality of robots 200a to 200n. In those embodiments, each electromagnetic signal 300 may have predetermined emission characteristics that allow identification of the robot of the plurality of robots 200a to 200n that emitted the electromagnetic signal 300.

[0197] In some embodiments, more than one of the plurality of robots 200a to 200n may each emit an electromagnetic signal 300 simultaneously to be detected by others of the plurality of robots 200a to 200n. In such scenarios, each electromagnetic signal 300 may have different emission characteristics (e.g. a different frequency slot in the same band) to the other simultaneously emitted electromagnetic signal(s) 300. The simultaneously emitted electromagnetic signals 300 may then interact with each other, and with objects or structures the environment 100, resulting in further distinctive sets of received characteristics that can be utilised in generating or populating the electromagnetic model 120 of the environment 100.

[0198] Returning to Figure 4, robots 200a, 200b, and 200c move around the environment 100 to generate or populate the electromagnetic model 120. Here, robot 200a acts as an "emitter", and emits an electromagnetic signal 300 that is detected by each of robots 200b and 200c. Robots 200b and 200c may detect the electromagnetic signal 300 either directly (i.e. after the signal has only propagated through the air), or after the electromagnetic signal 300 has been reflected, refracted, scattered off of, and/or transmitted through, for example, the various surfaces in the environment 100 (as in other figures, this is shown as dotted curved lines). As highlighted above, robot 200a may also detect electromagnetic signal 300, for example after it has been reflected, refracted, scattered off of, and/or transmitted through, for example, the various surfaces in the environment 100.

[0199] Fig ure 5 shows a flowchart of a method 2000 for the spatial intelligence gathering system 10 to generate or populate the electromagnetic model 120 of the environment 100 according to some embodiments of the present application.

[0200] It will be understood that any of the processes and functionalities discussed above may be applied in conjunction with, or as equivalents of, the steps of the flowchart shown in Figure 5. For example, in some embodiments the method steps of Figure 5 may be carried out by a single robot, rather than the plurality of robots 200a to 200n, in the manner discussed above.

[0201] Some of the steps shown in Figure 5 are optional steps (that is to say, not essential to work the invention of the present application). Such optional steps are indicated as such with a dotted line.

[0202] At Step 2100, a plurality of robots 200a to 200n are released into an environment 100, or may already be present in the environment 100. The plurality of robots 200a to 200n move around the environment 100 in a closed loop manner, where movement is directed through inter-communication between the one or more robots 200a to 200n_z and/or between the one or robots 200a to 200n and a server 250 of the spatial intelligence gathering system 10. Movement of each of the one or more robots 200a to 200n is further, or alternatively, directed based on the map of the physical layout 110 of the environment 100, and/or the current positions of each of the plurality of robots in the environment 100.

[0203] At Step 2200, one of the plurality of robots 200a to 200n emits an electromagnetic signal 300 into the environment 100. The emitted electromagnetic signal 300 includes a set of emission characteristics. The set of emission characteristics may include characteristics such as signal strength, frequency profile, duration, or any other suitable parameter of an electromagnetic signal.

[0204] The emitted electromagnetic signal 300 may include an identification parameter (e.g. an identification number, code, or other ID) for the robot that emitted the electromagnetic signal 300, as well as data indicating the location in the environment 100 at which the electromagnetic signal 300 was emitted by that robot (i.e. the emission location). The emitted electromagnetic signal 300 may additionally is include data indicating the time at which the electromagnetic signal 300 was emitted (i.e. the emission time).

[0205] Any of the identification parameter, emission location, and emission time may be part of the emitted electromagnetic signal 300, or may be transmitted or communicated to the other robots of the plurality of robots 200a to 200n separately to the electromagnetic signal 300. This may be achieved, for example, using the communications module 240 of the spatial intelligence node 50 of each robot. Any of the identification parameter, emission location, and emission time may be transmitted or communicated to the other robots of the plurality of robots 200a to 200n at the same time that the electromagnetic signal 300 is emitted, before the electromagnetic signal 300 is emitted, or after the electromagnetic signal 300 is emitted.

[0206] The electromagnetic signal 300 may have predetermined emission characteristics that allows for the identification of the robot of the plurality of robots 200a to 200n that emitted the electromagnetic signal 300.

[0207] At Step 2300, one or more of the other robots of the plurality of robots 200a to 200n may then detect (or receive) the electromagnetic signal 300 emitted by the one robot of the plurality of robots in Step 2200.

[0208] Each of the one or more of the other robots that detects (or receives) the electromagnetic signal 300 is located at a different location in the environment 100 to that of the one robot that emitted the electromagnetic signal 300 (i.e. the emission location) in Step 2200.

[0209] The electromagnetic signal 300 detected (or received) by each of the one or more of the other robots of the plurality of robots 200a to 200n has a set of received characteristics (e.g. signal strength, frequency profile, duration, or any other suitable parameter of an electromagnetic signal) for each robot that detects (or receives) the electromagnetic signal 300.

[0210] Each set of received characteristics of the electromagnetic signal 300 may be different to the set of emission characteristics of the emitted electromagnetic signal 300 as a consequence of the emitted electromagnetic signal 300 having propagated through the environment 100 from the emission location to the location of the robot of the one or more of the other robots of the plurality of robots 200a to 200n that detected (or received) the electromagnetic signal 300.

[0211] At optional Step 2350, each of the one or more of the other robots of the plurality of robots 200a to 200n that detected (or received) the electromagnetic signal 300 communicates or shares the set of received characteristics, as well as the location in the environment 100 at which the electromagnetic signal 300 was detected (or received) by that robot, with one or more of the other robots of the plurality of robots 200a to 200n (and/or, where included, the server 250). This may be achieved, for example, using the communications module 240 of the spatial intelligence node 50 of each robot.

[0212] At Step 2400, the spatial intelligence gathering system 10 generates or populates the electromagnetic model 120 of the environment 100 based on the set of emission characteristics compared with the sets of received characteristics for the electromagnetic signal 300.

[0213] At optional Step 2450, the method 2000 may return to Step 2200, where the same or a different robot of the plurality of robots 200a to 200n may emit a further electromagnetic signal 300. Here, the plurality of robots 200a to 200n may first move through the environment 100 to new locations before the further electromagnetic signal 300 is emitted, or may continue to move through the environment 100 and emit the further electromagnetic signal 300 after a period of time has elapsed since the last electromagnetic signal 300 has been emitted (e.g. a predetermined period of time) or in response to a triggering event (e.g. detecting a suitable location for such an emission, such as a cavity in a wall or a possible structural fault in a wall).

[0214] Throughout the above steps of method 2000, the plurality of robots 200a to 200n may move around the environment 100 and generate or populate the electromagnetic model 120 of the environment 100 by utilising SLAM techniques or any other suitable technique (e.g. using a global navigation satellite system, GNSS, or by using fiducial markers, ultrasound beacons, Bluetooth low energy beacons, etc.).

[0215] This approach of "active" generation or population of the electromagnetic model 120 of the environment 100 provides greater efficiency, reducing the time taken to populate the electromagnetic model 120 and producing an electromagnetic model 120 that is more accurate. In addition, the plurality of robots 200a to 200n may be directed around the environment 100 in manner that ensures that the locations at which the electromagnetic signals 300 are emitted are suitable locations for efficiently populating the electromagnetic model 120. For example, the locations for emitting the electromagnetic signal 300 may be chosen such that the electromagnetic model 120 is populated most efficiently for the given environment 100. Similarly, the locations of the one or more robots of the plurality of robots 200a to 200n that "listen" for the emitted electromagnetic signal 300 may be chosen in a manner such that the electromagnetic model 120 is populated most efficiently for the given environment 100. In view of the greater efficiency provided by this approach, it is also possible to reduce the number of robots used in the spatial intelligence gathering system 10 for this purpose. For example, the number of robots may be reduced to fewer than 10 robots when populating i the electromagnetic model 120 when using the "active" approach discussed above and shown in Figure 4.

[0216] The above "active" approach, as well as the other approaches discussed in the present application, can also be used to generate training data for machine learning algorithms, where such machine learning algorithms may be used to identify signals of interest in the environment 100. Here, different radio-frequency, RF, or EM fingerprinting and radio-domain/EM machine learning techniques are dependent upon access to complex training datasets, which can localise/overfit the machine learning algorithm and limit reuse. By comparison, the application of the spatial intelligence gathering system 10 described herein allows the plurality of robots of the system to build machine learning training datasets based upon their surroundings. This can then be used to build up a digital fingerprint for a location or building (i.e. an environment 100), allowing for more sensitive anomaly detection, or to track the structural health of a building with time.

[0217] Where the above "active" approach might be used to generate training data for machine learning algorithms for identifying signals of interest in the environment 100, the emitted electromagnetic signal 300 may be configured to imitate or simulate a potential signal of interest 70 in the environment 100, which may be characterised as such by the other one or more robots 200a to 200n (or, in some embodiments where a single robot is used as discussed above, that single robot) as part of generating the training data. The training data may then include a set of data reflecting the characteristics of various signals of interest, such that a machine learning algorithm may be trained using that training data to assess detected EM signals (in that environment 100 or in other environments) and characterise or categorise those detected EM signals as being signals of interest (or not as being signals of interest).

[0218] Such configuring of electromagnetic signals 300 to simulate signals of interest may include setting the signal characteristics of the electromagnetic signal 300 (e.g. signal strength, frequency, etc.) in a specific manner, and/or may include selecting a location for emitting the electromagnetic signal 300 which would usually cause a detected signal to be characterised as a signal of interest 70. For example, the electromagnetic signal 300 may be configured to simulate a signal of interest 70 by having the signal characteristics of a mobile phone, but the location may be selected to be very close to a light switch. As a further example, the electromagnetic signal 300 may be configured to simulate a signal of interest 70 by having the signal characteristics of the wireless connection of a desktop computer, but the location may be selected as being a wall (i.e. no observable electronic devices being present at all). [0219] It will be understood that whilst the above approach of Figures 4 and 5 are discussed with reference to developing an electromagnetic model 120 of an environment 100, this approach may also be applied to the ongoing monitoring of an environment 100 (e.g. a structure or building), when the electromagnetic model 120 has previously been generated. For example, the spatial intelligence gathering system 10 may employ such an "active" approach (i.e. emitting electromagnetic signals 300 from various locations in the environment 100) to monitor for changes in the electromagnetic behaviour of that environment 100 compared with a previously generated electromagnetic model 120 for that environment 100.

[0220] In addition, when one or more of a plurality of robots of the spatial intelligence gathering system 10 are used as "emitters" or "sounders" (i.e. the "active" approach discussed above with respect to Figures 4 and 5), this will generate a "fingerprint" even in the absence of electronic devices (i.e. a characteristic interaction between EM signals with given characteristics and the environment 100). This allows for the spatial intelligence gathering system 10 to be employed in otherwise empty, hazardous, or difficult to access environments, such as industrial plants, decommissioned nuclear power stations, buried/undersea infrastructure (such as tunnels, both underground and undersea).

[0221] Such a fingerprint for that structure may then be used to monitor for structural changes (e.g. structural fault development or subsequent degradation), since such changes would then be apparent from a change in the fingerprint over time (i.e. as the system 10 periodically monitors the structure). This can then be applied to monitoring the overall structural health of a structure or building, since such changes in electromagnetic behaviour (e.g. the propagation of EM signals) may indicate the development of structural defects in that structure, building, or site. Further examples include soil stabilisation monitoring and improvement, geophysical sensing, pipeline detection and condition assessment, tunnelling, trenching and trenchless technologies, structural performance of transport-ground-pipeline systems, and monitoring green/grey infrastructure interdependencies.

[0222] As discussed above, the systems, methods, and functionalities of the spatial intelligence gathering system 10 may be applied to a variety of different purposes (e.g. security and defence applications, electronic device monitoring, the characteristic interaction of a structure with the local radio frequency or electromagnetic environment, structural health monitoring, or archaeological surveying).

[0223] In some embodiments, the above functionalities (or their suitable equivalents), may be applied to sound sensing. In such scenarios, the electromagnetic model 120 may additionally or alternatively include modelling regarding the behaviour of sound in an environment 100 (i.e. being alternatively called a sound model 120 of the environment 100). In such scenarios, the one or more sensors 230 (and/or the at least one antenna 220) on the one or more robots may be selected to also detect and/or emit a wide range of different sound waves (e.g. ultrasonic ranging sensors, or other imaging sonar systems). As such, the spatial intelligence gathering system 10 may be configured to detect "sounds of interest" in an environment 100, and employ a sonar approach to "actively" map an environment 100.

[0224] In such scenarios, the spatial intelligence gathering system 10 may use such a sound-based approach to probe a structure or building (in an analogous manner to the use of EM signals discussed above) or be employed in a maritime context (e.g. for mapping underwater environments or identifying signals of interest underwater). It will nonetheless be appreciated that any of the methods or functionalities discussed herein with regard to EM signals may also be employed in environments that are underwater (or in any other medium other than air).

[0225] Where the spatial intelligence gathering system 10 as described herein comprises a plurality of robots 200a to 200n_z this may be described in terms of a "swarm" of such autonomous robots. Some or all of the individual robots in such "swarms" may be configured to operate to some degree independently, for example by individually generating a map of the physical layout 110 of the environment 100, or individually generating or populating an electromagnetic model 120 of the environment 100. As a result, multiple maps 110 and electromagnetic models 120 of the environment 100 may then exist within the system 10 at any given moment in time. Individual robots of the swarm may then make individual autonomous decisions based on their own unique maps 110 and electromagnetic models 120 (e.g. following a control algorithm, such as that discussed below). It will also be appreciated that the one or more robots 200a to 200n of a swarm, or the spatial intelligence gathering system 10 more generally, may send or receive data from other separate platforms in or near the environment 100 (e.g. a car, boat, plane, or other vehicle, manned or unmanned) that also utilise a spatial intelligence node equivalent to that included on one or more of the one or more robots 200a to 200n. This additional data may then contribute to, or be used in, any of the methods or functionalities discussed herein.

[0226] This allows for the various generated maps 110 and electromagnetic models 120 to be continually or periodically shared across the system 10, and combined, updated, reconciled, and/or reverted back to earlier versions depending on their performance or assessed accuracy. For example, if the closed loop behaviour of a robot is incorrect, or if a robot is found to have faulty sensors, then the map 110 and/or electromagnetic model 120 of that robot may be corrected or reverted using the map 110 and or electromagnetic model 120 of another robot(s). In such scenarios, the system 10 may keep a central "master" version of the map 110 and/or electromagnetic model 120 of the current environment 100 (e.g. stored by a "lead" robot or in the server 250, where included), where such master versions are combined and optimised from the map 110 and/or electromagnetic model 120 of various robots in the swarm. Such master versions may be periodically updated as the system 10 identifies objective improvements in the individual maps 110 and/or electromagnetic models 120 of various robots in the swarm. Similarly, individual maps 110 and/or electromagnetic models 120 of various robots in the swarm may be updated based on the master versions of the system 10. Individual robots in the swarm may also periodically share sensor data (such as from sensors 230) with each other and (where included) with the server 250, which can then be used by other robots in generating their individual maps 110 and/or electromagnetic models 120 of the environment 100.

[0227] In addition to general advantages in terms of adaptability and flexibility, this "swarm" approach is particularly advantageous in environments 100 where the robots in a swarm are distributed across a wide area and only communicate intermittently (e.g. where robots only periodically come within communication range of each other or of a common base station). In particular, each robot of the swarm can largely operate independently of the others, but can nonetheless still periodically benefit from the updated map 110, electromagnetic model 120, sensor data, or other data shared by other robots when in communication range.

[0228] The spatial intelligence gathering system 10 discussed herein may be utilised as an additional functionality in existing autonomous robotic systems. For example, the spatial intelligence gathering system 10 may be employed as an additional functionality for one or more robots also employed in a patrolling or security context, or as a logistics or cleaning robot. In such scenarios, the one or more robots may be configured to pause or suspend one functionality in order to investigate a signal of interest 70 in its current environment.

[0229] It will be understood that in the various embodiments discussed herein, the controlling or directing of the movement of the one or more robots 200a to 200n around the environment 100 may in some cases be said to be determined by a "control algorithm" 400. Here, such a control algorithm 400 may encompass any of the principles discussed in the embodiments herein in order to achieve the desired outcome. For example, the map of the physical layout 110 and the electromagnetic model 120 of the environment 100 may be utilised by a control algorithm 400 to direct the movement of one or more robots 200a to 200n around the environment 100 (for example to identify the source of a signal of interest 70). Inter-robot communication, and/or communication between the one or more robots 200a to 200n and a server 250 (where included), may be used by the control algorithm 400 to enable this (e.g. by sharing location and sensor data between robots and/or the server 250).

[0230] Fig ure 6 shows a representation of a control algorithm 400 employed in this manner, where one or more of detected EM signals (when generating or populating the electromagnetic model 120), current spatial data from sensors 230, the electromagnetic model 120, the map of the physical layout of the environment 100, the current positions and movement waypoints of other robots, and/or the detected signal of interest 70 (when identifying the source of a signal of interest 70) may be input into the control algorithm 400. The control algorithm 400 may then output appropriate movement waypoints or movement instructions for the robot in question based on those inputs. Alternatively, or in addition, the control algorithm 400 may employ such inputs to generate and output emission instructions, for example when an "active" approach is taken to generate or populate the electromagnetic model 120. Alternatively, or in addition, the control algorithm 400 may employ such inputs to generate and output antenna instructions for the emission and reception of electromagnetic signals, for example such that the at least one antenna 220 of a robot may be directed in a chosen direction for the emission or reception of electromagnetic signals. This may then be implemented using a horn antenna, beam steering, or beam forming (e.g. where the at least one antenna 220 is an antenna array), or where the antenna direction is variable through the use of an actuator or motor, or by any other suitable approach.

[0231] It will be understood that other inputs may be included in the control algorithm 400, as necessary to produce appropriate movement waypoints or movement instructions (and/or, where appropriate, EM emission instructions) (e.g. the most efficient for completing the current task) for the robots in question. As such, and where included, the control algorithm 400 enables the coordination of robot movements (and, where appropriate, EM signal emission), in order to ensure efficient and effective environment mapping (either physical mapping or electromagnetic modelling) or signal of interest 70 detection and investigation.

[0232] Such a control algorithm 400 may be implemented by a "lead" robot of the one or more robots 200a to 200n (e.g. by the spatial intelligence node 50 of that lead robot), may be implemented by the server 250 (where included), may be implemented separately by each robot of the one or more robots 200a to 200n_z or by any other part of the spatial intelligence gathering system 10.

[0233] Any of the above discussed methods may be performed using a computer system or similar computational resource, or system comprising one or more processors and a non-transitory memory storing one or more programs configured to execute the is method. Likewise, a non-transitory computer readable storage medium may store one or more programs that comprise instructions that, when executed, carry out the methods described herein.

[0234] Whilst certain embodiments have been described, these embodiments have been presented by way of example only, and are not intended to limit the scope of the application. Indeed, the novel devices, and methods described herein may be embodied in a variety of other forms; furthermore, various omissions, substitutions and changes in the form of the devices, methods and products described herein may be made without departing from the scope of the present application. The word "comprising" can mean "including" or "consisting of" and therefore does not exclude the presence of elements or steps other than those listed in any claim or the specification as a whole. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. The accompanying claims and their equivalents are intended to cover such forms or modifications as would fall within the scope of the application.

Claims

1. A method for mapping electromagnetic signals by one or more autonomous robots in an environment, the method comprising: mapping the physical layout of the environment; generating an electromagnetic model of the environment; detecting, by at least one of the one or more robots, an electromagnetic signal in the environment; characterising the detected electromagnetic signal as a signal of interest; directing the movement around the environment of at least one robot of the one or more robots based on the signal of interest; and identifying the source of the signal of interest in the environment.

2. The method of claim 1 wherein generating the electromagnetic model of the environment is based on the mapping of the physical layout of the environment.

3. The method of claim 1 or 2, wherein directing the movement of at least one robot of the one or more robots based on the signal of interest comprises directing the at least one robot in a manner to identify the source of the signal of interest.

4. The method of claim 3, wherein directing at least one robot in a manner to identify the source of the signal of interest comprises the robot moving based on one or more of the characteristics of the signal of interest, the electromagnetic model of the environment, and/or the mapping of the physical layout of the environment.

5. The method of any preceding claim, wherein once the source of the signal of interest is identified, directing the movement of the at least one robot of the one or more robots based upon further detected electromagnetic signals in the environment.

6. The method of any preceding claim, wherein the mapping the physical layout of the environment is based on one or both of: at least one robot of the one or more robots collecting sensor data from the environment surrounding that robot; or ii retrieving data on the physical layout of the environment stored in a memory device.

7. The method of claim 6, wherein when the at least one robot of the one or more robots collects sensor data from the environment, one or more of the mapping of the physical layout of the environment, the generating of the electromagnetic model of the environment, or the directing of the movement around the environment of at least one robot of the one or more robots based on the signal of interest comprises directing the one or more robots in a closed loop manner.

8. The method of claim 6 or 7, wherein when the at least one robot of the one or more robots collects sensor data from the environment, that at least one robot employs simultaneous localization and mapping techniques to map the physical layout of the environment and/or to generate the electromagnetic model of the environment.

9. The method of any preceding claim, wherein the one or more robots is a plurality of robots, and wherein the method further comprises: each robot of the plurality of robots communicating with another robot of the plurality of robots in order to carry out the steps of at least one of: mapping the physical layout of the environment; generating the electromagnetic model of the environment; directing the movement around the environment of at least one robot of the plurality of robots based on the signal of interest.

10. The method of any preceding claim, wherein characterising the detected electromagnetic signal as a signal of interest comprises at least one of: determining that a characteristic of the detected electromagnetic signal is not an expected characteristic for electromagnetic signals for the environment; determining that the detected electromagnetic signal has an unexpected emission pattern over time; determining an approximate location of the source of the detected electromagnetic signal and determining that the determined approximate location is not an expected location for the detected electromagnetic signal; determining a device as being a possible source of the detected electromagnetic signal and determining that the device is not expected to be a source of the detected electromagnetic signal; and/or determining that the detected electromagnetic signal has not been detected before.

11. The method of any preceding claim, wherein the generating of the electromagnetic model of the environment comprises: emitting a first electromagnetic signal into the environment from an emitting location in the environment, the first electromagnetic signal having a set of emission characteristics; detecting, by at least one of the one or more robots, the first electromagnetic signal, the at least one of the one or more robots being at a respective receiving location within the environment, wherein the first electromagnetic signal detected by the at least one of the one or more robots has a corresponding set of received characteristics; modelling the propagation of the first electromagnetic signal based on the emitting location and the set of emission characteristics compared with each of the receiving locations and the corresponding sets of received characteristics.

12. The method of claim 11, wherein the emitting of the first electromagnetic signal is by one of: at least one robot of the one or more robots; a static emitter located at the emitting location in the environment; a non-autonomous mobile platform; or a mobile hand-held emitter at the emitting location in the environment.

13. The method of claim 12, wherein the emitting of the first electromagnetic signal is by at least one robot of the one or more robots, and wherein the detecting of the first electromagnetic signal comprises the same at least one robot of the one or more robots detecting the first electromagnetic signal.

14. The method of any preceding claim, wherein the one or more robots is a plurality of robots, and wherein : directing the movement of at least one robot comprises directing more than one of the plurality of robots based on the signal of interest; and ii the plurality of robots communicate with each other to identify the location of the source of the signal of interest within the environment.

15. The method of any preceding claim, wherein the one or more robots is a plurality of robots, and wherein determining that the detected electromagnetic signal is a signal of interest comprises a machine learning algorithm determining that the detected electromagnetic signal is a signal of interest; and wherein training data for the machine learning algorithm comprises data corresponding to at least one of the plurality of robots emitting a simulated signal of interest from a location in the environment, and at least one of the plurality of robots detecting the simulated signal of interest and identifying the location as the source of the simulated signal of interest.

16. The method of any of claims 11 to 13, wherein the emitting location is a location in the environment at which a signal of interest might be expected to originate.

17. A system for modelling electromagnetic signals in an environment, the system comprising: one or more autonomous robots, each robot of the one or more robots comprising: at least one sensor for collecting data such that the robot can navigate the environment; at least one sensor for collecting data relating to electromagnetic signals in the environment; and a communications module for transmitting the collected sensor data; wherein the system is configured to carry out the method of any of claims 1 to 16.

18. The system of claim 17, wherein one or more of the one or more robots further comprises a spatial intelligence node, the spatial intelligence node comprising: at least one sensor for localising the one or more of the one or more robots in the environment; and at least one antenna for emitting and receiving electromagnetic signals.

19. A non-transitory memory storing one or more programs, which, when executed by one or more processors of a device, cause the device to perform the method of any one of claims 1 to 16.