BACKGROUND

The tracking and monitoring of athletes in real-time is of great interest to players, coaches, broadcasters and fans. In the past, this has been achieved using a combination of a Global Navigation Satellite System (GNSS) and a wireless protocol, such as Bluetooth (RTM) or Wi-Fi (RTM). More recently there has been a shift to ultra-wideband (UWB) systems, which can accurately locate players and equipment in real-time and in such a way that the data is not affected by the noisy radio frequency (RF) environment that is often present in crowded stadia.

Typical UWB systems use a combination of tags (electronic devices which are mobile and which are actively tracked) and anchors (electronic devices which are fixed in known positions and form a reference system, relative to which the tags are located). The tags are worn by athletes, or are embedded within sports equipment, and are thus representative of movement of the athlete or the equipment.

Tracking systems have been described which combine GNSS and UWB technologies. For example, in the tracking of drones, GNSS and UWB technologies are used in conjunction with one another to enable the drone to navigate both indoors within a UWB infrastructure and outdoors where GNSS takes over. In the field of sports tracking, UWB has been proposed as an augmentation or alternative to GNSS, making use of fixed anchors placed around the field of play. For example, WO 2017/174956 proposes a system where position is calculated using both a GNSS technique and an UWB technique and the best candidate position is chosen. It has also been proposed to use GNSS to identify the location of the fixed pitch-side anchors and to calculate the position of the tags using UWB relative to those fixed, stationary anchors.

GNSS-led solutions are generally undesirable as performing regular position measurements also requires significant battery power which makes electronic devices heavy and impractical for use in high-performance events. The use of a set of stationary anchors for UWB ranging works well in many scenarios. If the positions of the anchors are accurately known, then the mobile tags worn by the players may be located to a similarly high accuracy. However, there are scenarios where using fixed anchors is inconvenient. Installation of these fixed anchors is time-consuming, expensive and a source of irritation for potential users, since the players or equipment can only be tracked at the location where the anchors are installed, for example stadia or defined training pitches.

Stationary anchors are typically placed at high elevations to improve signal accuracy, for example they may be affixed to lighting poles or to the roof of a venue. This brings with it significant additional challenges. Mounting the anchors high-up is dangerous for the installation technicians and so the installation takes longer and is more expensive. Certain venues do not have convenient infrastructure on which to mount the anchors, and at these venues large tripods must be used, or long poles installed into the ground. This additional infrastructure adds to the cost of the installation.

The reliance of pitch-side infrastructure such as stationary, fixed anchors in real- time sports tracking significantly limits the usage and adoption of such systems.

SUMMARY OF THE INVENTION

The present invention provides a system and method for real-time object tracking using UWB and GNSS without the need for fixed anchors.

According to an aspect of the present invention there is provided a method of identifying position of moving electronic devices on a sports field, the method comprising: calculating a linear distance between each of at least three moving electronic devices and at least two other moving electronic devices of the at least three moving electronic devices using an ultra-wideband radio signal to create a set of linear distance measurements; creating a rigid device topology of the plurality of moving electronic devices using the set of linear distance measurements; retrieving a position measurement of at least three of the at least three moving electronic devices; identifying a position of each electronic device by determining arrangement of the device topology using the position measurements.

In this context, each refers to every one of two or more people or things regarded and identified separately. The two other moving electronic devices may for example be neighbours in the topology.

Accordingly, a solution is provided in which moving devices range to each other to define a topology of the system. This topology is floating until fixed by positional information of at least some of the devices. Not all devices need to know their position but they must need only their relative location in the topology.

Accordingly, the positions of moving electronic devices can be located to a high accuracy without the inconvenience of time-consuming, expensive and dangerous installation. The system can function indoors and can be moved to new pitches or environments without a new installation. Accuracy problems caused by line of sight challenges are obviated.

The rigid topology may for example be a rigid polytope where an amount of linear measurements between nodes are identified in order to make the topology rigid. For each rigid polytope, preferably at least 3 positional measurements may be determined to fix that rigid polytope, even if the polytope includes nodes which are not located by the positional measurement but are only determined relative to other nodes in the polytope, to make it rigid. Devices not forming part of the rigid polytope, i.e. floating, may be fixed with a single positional measurement and linear measurements to other nodes of the polytope (even if not enough linear measurements to make the device part of the rigid polytope).

The step of retrieving a position measurement comprises retrieving a GNSS position measurement of at least three of the at least three moving electronic devices. In this way, a technique is provided whereby UWB and GNSS can be used to locate athletes and sports equipment in real-time without the need for fixed anchors. The UWB is not used to make the GNSS more accurate, rather the GNSS is used to fix the position of the UWB calculations, or at least approximate the position of the UWB calculations. In this solution there is limited differentiation between an anchor, that is pitch-side infrastructure, and a tag, that is an object to be tracked. The tags themselves function as anchors and determine location relative to one another.

Moreover, the power consuming GNSS measurements need not be obtained with the same regularity as the UWB measurements. Once the topology of the nodes is known, the accurate UWB measurements can be used to identify distances or locations and more particularly distances changes. Thus, the GNSS measurements may be made less frequently while maintaining accuracy of the system.

The step of retrieving a position measurement comprises: calculating a linear distance between each of at least three of the at least three moving electronic devices and three stationary electronic devices using an ultra-wideband radio signal, wherein the three stationary electronic devices have a known position in a coordinate system. The three stationary electronic devices may be transitory and configured to be positioned around the periphery of a sports field.

This provides for accurate determination of the position of a moving electronic device and also the position of a mobile device which is not in communication with the stationary anchors. High power consumption GNSS measurements need not be obtained and thus the tags can be made small, lightweight and long- lasting which is preferable when these tags are worn by athletes and is essential when they are inserted in a ball. Similarly, the system can be used where GNSS systems have been shown to be inaccurate such as indoors or in stadium environments (or other inferring environments). The non-permanent anchors need not be located using survey equipment or located at height as they do not provide the accurate location measurements of each tag. The step of identifying a position of each moving electronic device may comprise identifying the position of each electronic device in a coordinate system by determining arrangement of the device topology in the coordinate system. The coordinate system may be a global coordinate system.

Previous work in the field uses UWB to make the GNSS more accurate. The invention proposes an infrastructure free UWB tracking solution which uses positional measurements through either GNSS or transitory pitch-side anchors to fix the floating UWB coordinate system. The tag-to-tag UWB is not used to make the GNSS or anchor UWB more accurate, rather the positional measurements (e.g. GNSS or anchor UWB) are used to fix a floating UWB coordinate system to a global coordinate system.

The step of identifying a position of each electronic device may comprise using the position measurement to determine a transformation of the device topology which fixes the electronic devices relative to earth.

The step of identifying a position of each electronic device may comprise determining rotation, translation and reflection parameters of the topology from the position measurement. The UWB part of the system measures the relative displacements of the athletes and their sports equipment. The locations of the athletes and sports equipment may then be determined from the displacements determined through the UWB measurements, however, with a rotational, translational and reflection uncertainty. Position measurements (from GNSS or anchor UWB) are then used to determine these unknown rotation, translation and reflection parameters. Again, only a few position measurements need to be identified to fix entire topology of devices, that is, devices need only located in the topology which itself is fixed. Again, preferably that topology is rigid.

Preferably the method further comprises determining arrangement of the device topology using an estimation algorithm configured to estimate rotation, translation and reflection parameters which best fit the position measurements (e.g. GNSS). More preferably the estimation algorithm includes a least-squares model. The inaccuracy of the position measurements can be compensated for using the accurate UWB measurements and the UWB measurement topology of which the accuracy can be confidently assumed is made to best fit the position measurements by solving for the rotation, translation and reflection parameters to approximate the topology to the position measurements.

Preferably the moving electronic devices are attached to objects on the sports field. For example the devices may be embedded within a ball or strapped to or otherwise worn by the player. Accordingly, the technique can track the movement and position of objects on the field of play.

Calculating a linear distance using an ultra-wideband radio signal may comprise determining the linear distance using two-way ranging. The devices can thus accurately and in real-time calculate the distance between each other using the time taken for radio signals to travel between them.

At least one of the electronic devices may act as a master clock and each of the other electronic devices synchronise in time from the master clock. Thus the master clock is used to synchronise all the devices to the same time using a timing network.

In one example, the electronic devices synchronise in time using the two-way ranging message. Thus the time is efficiently distributed..

Optionally, synchronising the electronic devices in time further comprises receiving a clock synchronisation signal from a different electronic device. For example, the ball or a pitch-side device may act to distribute a master clock to synchronise all the player devices to enable accurate measurements.

In one example, the two-way ranging can be symmetric. Thus delay and error can be mitigated.

At least one of the electronic devices acts as a master device, wherein the method comprises, at the master device: identifying a predetermined list of devices; and, initiating a linear distance calculation to the electronic devices by sending a messages to the others of the electronic devices. Thus the process of calculating distance can be easily managed and controlled by a master device. The list of devices may for example be programmed into the system before the game begins and the devices are embedded within the sports equipment.

In certain embodiments, the method may further comprise, at each moving electronic device of the at least three moving electronic devices, calculating the linear distance between the respective electronic device and at least two of the other moving electronic devices and sending the linear distances to a data processing apparatus. Thus the distances are offloaded to a data processing apparatus which is able to track and identify the positions of all of the devices on the field from the distance measurements.

According to a further aspect of the invention there is provided a system for identifying position of moving electronic devices on a sports field, the system comprising: at least three moving electronic devices configured to be embedded within sports equipment on a sports field, each moving electronic device comprising an device antenna connected to a respective device transceiver, wherein each moving electronic device is configured to: calculate a linear distance between the respective electronic device and at least two others of the moving electronic devices using an ultra-wideband radio signal; and, send the linear distances to a remote processing unit, and wherein at least three of the at least three moving electronic devices are configured to: generate a position measurement of the respective electronic device; and, the remote processing unit configured to: receive the linear distances and the position measurements; create a rigid device topology of the plurality of moving electronic devices using the set of linear distance measurements; and, identify a position of each moving electronic device by determining arrangement of the device topology using the position measurements.

At least three of the at least three moving electronic devices may be configured to generate a position measurement by identifying a GNSS position measurement of the respective electronic device. Alternatively or additionally, the system may further comprise at least three stationary electronic devices, and wherein at least three of the at least three moving electronic devices are configured to generate a position measurement by calculating a linear distance between the respective moving electronic device and the three stationary electronic devices using an ultra-wideband radio signal.

The three stationary electronic devices may be transitory and configured to be positioned around the periphery of a sports field.

The remote processing unit may be configured to identify a position of each moving electronic device by identifying the position of each moving electronic device in a coordinate system by determining arrangement of the device topology in the coordinate system. The coordinate system may be a global coordinate system.

The remote processing unit may be configured to identify a position of each moving electronic device using the GNSS position measurement to determine a transformation of the device topology which fixes the moving electronic devices relative to earth.

The remote processing unit may be configured to identify a position of each electronic device comprises determining rotation, translation and reflection parameters of the topology from the GNSS measurement.

The remote processing may be configured to determine arrangement of the device topology using an estimation algorithm configured to estimate rotation, translation and reflection parameters which best fit the position measurements. The estimation algorithm may include a least-squares model.

The moving electronic devices may be attached to objects on the sports field. The electronic devices may be configured to calculate a linear distance using two-way ranging. The two-way ranging may be symmetric. At least one of the electronic devices may be configured as a master clock and each of the other electronic devices is configured to synchronise in time from the master clock. The electronic devices are configured to synchronise in time using the two-way ranging message. At least one of the electronic devices may be configured to synchronise in time by receiving a clock synchronisation signal from a different electronic device. At least one of the electronic devices may be configured to act as a master device and may be configured to: identify a predetermined list of devices; and, initiate a linear distance calculation to the electronic devices by sending a message to the electronic devices.

The moving electronic devices may be configured to calculate the linear distance between the respective electronic device and at least two of the other electronic devices and send the linear distances to a data processing apparatus.

According to a further aspect of the invention there may be provided a computer readable medium comprising instructions which, when executed by a processor, cause the processor to perform the method of the first aspect.

According to an aspect of the present invention there is provided a method of identifying position of moving electronic devices on a sports field, the method comprising: calculating a linear distance between each of at least three electronic devices using an ultra-wideband radio signal to create a set of linear distance measurements; creating a device topology of the plurality of electronic devices using the set of linear distance measurements; retrieving a GNSS position measurement of at least three of the at least three electronic devices; identifying a position of each electronic device by determining arrangement of the device topology using the GNSS position measurements.

According to a further aspect of the invention there may be provided a system for identifying position of moving electronic devices on a sports field, the system comprising: at least three electronic devices configured to be embedded within sports equipment on a sports field, each device comprising an device antenna connected to a respective device transceiver, wherein each electronic device is configured to: calculate a linear distance between the respective electronic device and at least two others of the electronic devices using an ultra-wideband radio signal; identify a GNSS position measurement of the respective electronic device; and, send the linear distances and the GNSS position measurement to a remote processing unit; and, the remote processing unit configured to: receive the linear distances and the GNSS position measurements; create a device topology of the plurality of electronic devices using the set of linear distance measurements; and, identify a position of each electronic device by determining arrangement of the device topology using the GNSS position measurements.

DETAILED DESCRIPTION

Examples of systems and methods in accordance with the invention will now be described with reference to the accompanying drawings, in which:-

Figure 1 shows a high-level schematic diagram of a prior art system;

Figure 2 shows a schematic diagram of four tags and their linear distances; Figure 3 shows a schematic diagram of a topology of four tags and a possible configuration;

Figure 4 shows a schematic diagram of a topology of four tags and a possible configuration on the field;

Figure 5 shows a schematic diagram of a topology of four tags and possible GNSS measurements;

Figure 6 shows a flow diagram of a process according to an example of the present invention; and,

Figure 7 shows a schematic system diagram of a system according to an example of the present invention

Figure 8 shows a flow diagram of a process according to an example of the present invention; and,

Figure 9 shows a schematic diagram of a topology of four tags, three anchors and a possible configuration on the field.

The following are examples of systems and methods for tracking and monitoring of movable elements, such as players and balls, in complex sports such as Rugby Football. The principles are also applicable to American Football, Soccer and other sports. It will be understood of course that the following are merely examples. The principles described are usable for performance analysis and training purposes as well as officiating and broadcasting. More specifically the exemplary concepts relate to a set of wearable devices on the players and devices embedded into the ball, or other equipment, which enable real-time detections to be made with high accuracy and precision without the need for fixed infrastructure.

Known state of the art systems for location detection in sports include those disclosed in GB 2,541 ,265, which is hereby incorporated by reference. In this example system, wearable devices use ultra-wideband transceivers to communicate with devices at the side of the field which calculate the position and orientation of the devices from the signals received.

More widely used systems include data loggers which utilise the Global Navigation Satellite System (GNSS) as well as accelerometers and gyro sensors to track position and record the speed and position of a player as they move around a pitch. The position of the player and their wearable device is calculated using a series of satellites or nodes positioned around the playing area which communicate with the device using 2.4GHz wireless communication, such as Wi-Fi (RTM) or Bluetooth (RTM).

Figure 1 illustrates a system of the former example of the state of the art. In this example system, wearable devices use a combination of ultra-wide band (UWB) transceivers. These devices make it possible to determine of the location, velocity, acceleration and angular orientation of the player while measuring the direction, magnitude and location of the dynamic forces experienced by the player. An UWB transceiver and a 9-axis Inertial Measurement Unit (IMU) may be embedded within the rugby ball to allow for the location, velocity, acceleration and angular velocity components of the ball to be determined. These devices may broadcast their data in real-time to apparatus at the side of the field or offload the data after the recording activity. Known figure 1 illustrates a rugby pitch 10 on which is positioned an article of body armour 11. Positioned around the pitch 10 are a series of antenna arrays or anchors 13. Figure 1 illustrates a device embedded within body armour to be worn by a player, as is typical in Rugby Football. A ball may also include a similar device embedded within its interior. A central server (not shown) is used to process the data channelled from the players and equipment via the receivers.

Embedded within the body armour 1 1 in this known example is an electronic device 12 which includes an antenna and one or more ultra-wide band (UWB) transceivers. A similar device may also be embedded in other equipment such the ball or training equipment. The UWB transceivers in the electronics device transmit a narrow pulse in the time domain, or‘chirp’, which is detected by the anchors at different times. The anchors perform Angle of Arrival (AoA), Time Difference of Arrival (TDoA) and Time of Arrival (ToA) calculations to determine the player’s position. The principles behind AoA, TDoA and ToA are well known and will not be described in detail here. Similarly, the principles behind two-way ranging, symmetric two-way ranging and asymmetric two-way ranging which may be used are well known and well documented.

In the example of the art illustrated, to determine position, multiple signals can be compared to trilaterate and/or multilaterate the position once the relative positions of the fixed anchors are known. Each anchor may be connected to the others wirelessly or through fixed or wired communications. Each anchor may also be connected solely to a master anchor which gathers together data from each anchor and instructs each anchor to act. The master anchor may be connected to a server or the server may act as the master and coordinate the anchors.

The master anchor may also function to set the reference coordinates to (x,y,z) (0,0,0). Thus the reference frame against which the orientation of the device is determined is set by the master anchor. The coordinates of the reference frame are then based on the position of other anchors. In order to determine the reference frame for position and orientation, more than one point is needed. Each anchor position is determined relative to each other and then positioned on the frame, all relative to the master. The position and orientation of the device can therefore be considered to be relative to the frame set by the anchors. The origin may be placed anywhere on or around the pitch, but preferably the y-axis is placed at the halfway line. The other side of the halfway line would therefore be, for example, (0, 70, 0).

The terms tag, electronic device, Sports Monitoring Device (SMD), wearable, and embedded device are used interchangeably throughout the present application. Similarly, as will be understood, the terms beacons, radio-beacons, anchors and antenna arrays signify the apparatus installed around the periphery of the field and are used interchangeably.

As indicated above, player tracking has been typically been accomplished predominantly through the use of GNSS (for example GPS or Glonass). GNSS is a convenient, cost effective solution for player tracking and in isolated, outdoor environments can be relatively accurate. Since the satellites are installed, paid for and maintained by third parties, the solutions can be cost effective. Nevertheless, in stadium and indoor environments, the solutions are inaccurate and in some cases do not work at all.

Alternative UWB systems are relatively new and their adoption has been slow because of high cost of the electronics and the high cost of the installation. Another impediment to the adoption of UWB systems is that they are only useful for locating objects on the field around which they were installed. For example, if a football team trains on a different pitch to the one they play on, then two UWB systems will be needed.

As mentioned already, in the present description a system is proposed for tracking tags accurately using UWB without the need for anchors. That is, without the need for fixed infrastructure of which the position is known to a high degree of accuracy. The process determines a possible topology from a set of relative linear distances between tags and then identifies the most likely arrangement of that topology. Preferably the process accurately determines the topology of the tags using UWB and then fixes that topology to a coordinate system using GNSS.

Figure 2 schematically illustrates four tags. The process by which tags may first identify their linear distance is well known in the art. A preferred technique is ultra-wideband two-way ranging or, optionally, symmetric two-way ranging. Since there are no anchors the tags identify their position relative to one another rather than relative to the anchors. Accordingly each tag may be thought of to function as an anchor as well as a tag in standard two-way ranging parlance. Alternatively, the system may be thought of as having no anchors at all and only tags.

Since ranging techniques are well known and well documented, we do not describe them in detail here but for context we provide a high level summary. A first device A may send a poll message to a second device B. An acknowledgement of that poll is returned by the second device B. A final message is sent from the first device A in reply to the acknowledgement. From these three messages the second device B can identify its distance to the first device A. Each of the three messages may contain the time at which it was sent. By identifying the time at which the poll message received, the second device B can calculate the signal propagation delay of the poll message. The second device also knows the processing delay representative of the time it takes the second device to process the poll, generate the acknowledgement and transmit it. A further propagation delay is identified which is the time it takes for the acknowledgement message to return to the first device A. This time (or delay) is included in the final message from the first device A to the second device B. Using these three pieces of information, the second device B can accurately identify its linear distance from the first device A. Optionally, a verify message may be sent from the second device B to the first device A after the reply message so that the first device A can check the measurements and accurately determine its distance taking into account clock or other errors. At least one of the tags may act as a master clock. The remaining tags may synchronise in time with this master clock. Techniques for compensating for clock drift are well known in the art. For example, if a message is sent from the master clock tag at regular intervals, the secondary tag can time these messages relative to their own clocks. If these messages are identified at unexpected times then the clock may be inaccurate relative to the master. Not all tags need to communicate with the master to synchronise clocks, as the compensation can be propagated from tag to tag.

Optionally, a different device from the tags may act as the master to synchronise clocks and/or organise the timing network. For example, a pitch-side device or a device embedded within the ball which is different from the player tags.

The techniques for compensating for clock drift may be part of a timing network. Such timing networks are well known. The timing network is used to arrange the communication of tags. Each tag may be assigned a seat or timeslot within which they may communicate to reduce interference. Accordingly the timing network may be used to assign each of the tags a time period within a time wider period during which distance measurements may be made to the other tags. The master tag may optionally arrange this orderly signalling.

The orderly scheme of communication allows the tags to communicate with each of the other tags to identify the linear distance between each tag.

In a preferred implementation of the invention, each tag may be assigned a predetermined seat prior to the process beginning and this may be given to the master tag to coordinate the position measurements. Alternatively a master tag may wake up and broadcast a message asking for replies from tags in communication range and the master may generate the list of tags and seats using a discovery process, that is, the master may optionally assign to the available tags a timeslot within which to send and receive its distance measurement messages. As illustrated schematically in Figure 2, four example tags 21 , 22, 23, 24 each perform distance measurements to, or with, the others and each may store the distance data associated with a timestamp of the time that measurement was made in a datalogger internal to the device.

These timestamped distance measurements may be offloaded to a remote server in real-time or offloaded in a batch. The specific method of data offloading to a server is not important to the innovation. The following describes the algorithm performed at a remote server on saved measurements however it is equally contemplated that the location determined may be performed on one or more of the tags themselves and the accurately determined positions stored or sent to a remote location.

From these timestamped distance measurements, the server is able to determine a topology of the tags. The topology is effectively equivalent to a mesh. Consistent with such network topologies, each tag or node does not need to identify the linear distance to every other node to accurately identify the complete topology or configuration of tags on the field. Each tag will identify the relative linear distance between the tag and at least two of its neighbours, preferably at least three. In the given example there are four tags and three linear distance measurements. The topology may be thought of here as rigid and thought of as a rigid polytope. A polyhedron is rigid if it cannot be continuously deformed into another configuration. A rigid polyhedron may have two or more stable forms which cannot be continuously deformed into each other without bending or tearing.

These UWB distance measurements give only the mutual distance of the players and/or sports objects, as shown in Figure 2. This configuration can be rotated, flipped or translated elsewhere and the mutual distances will remain unchanged. Therefore, if fixed anchors are not used, the UWB measurements from the freely moving athletes and/or sports objects, define infinitely many configurations which differ by a linear combination of a translation, rotation and reflection, as shown in Figure 3. That is, there are infinitely many alternative configurations of the topology which exist which have the same distance measurements. In this figure it is shown that the configuration may be altered without varying the mutual distances between the tags and their topology.

In the topology these distances can optionally be thought of also as mutual displacements. That is, in the topology the scalar distances determined from the two-way ranging is made a vector using the measurements made between the other nodes.

The topology can also be considered a connected graph. In the graph the players or balls are the nodes and the relative distances are the edges. The graph is well defined in shape— one can imagine holding a model in a hand and walking around with it, rotating it etc. That is, the graph is not set in place as the arrangement of tha graph cannot be determined by the algorithm. This metaphor is a good image because it is aids the understanding of how the rigid topology (or connected graph) is well-defined in shape and feel but indeterminate in location, rotation and reflection. Here rigid refers to the fact that the conceptual polyhedron can be rotated but not bended. Figure 3 illustrates how, without fixed anchors, UWB distance measurements between players define infinitely many valid configurations which differ by a linear combination of a translation, rotation and reflection.

Figure 4 demonstrates the effect of this unknown configuration on a sports field. In both two and three dimensions, the specific position of the tags cannot be determined accurately because the valid configuration of the topology cannot be determined. By determining the topology first, the relative distances between the tags can be determined accurately as UWB measurements are highly accurate and in most practical circumstances do not suffer significantly from interference. However, the accurate location of the tags cannot be determined from this topology information.

The proposed process continues by identifying a GNSS position of at least three of the tags in the topology. Based on these GNSS measurements the process is able to fix the device topology. Given that there are infinitely many configurations which satisfy the distance measurements, the GNSS position data is subsequently used to determine the transformation which fixes the athletes and/or equipment, preferably relative to the earth. This invention therefore allows for accurate location of athletes and/or equipment without having to install fixed infrastructure.

In a preferred implementation, a GNSS measurement is stored by at least three of the tags alongside the relative distance measurements made at a specific time and associated with the same timestamp in the datalogger. The data is either used by the tag for calculation or offloaded as described above.

Preferably the implementation of the process then applies an estimation algorithm to the topology data stored and the three GNSS measurements. For example, the estimation algorithm may be a least squares or least square error algorithm to fit the data. The algorithm may solve for a possible translation, rotation and flip of the topology that most likely fits the GNSS measurements. Many different estimation, adaption or approximation algorithms are contemplated as would be understood by the skilled person. The preferred approach is a least-squares model, which solves for the most approximate translation, rotation and reflection parameters. Any suitable method for estimating the most likely combination of parameters to fit the GNSS measurements is contemplated here. Recall that the topology is considered accurate in most circumstances and subsequently fit to inaccurate GNSS measurements in the best or most likely way.

The algorithm can be considered to apply a sequence of linear transformations and minimise the variables be minimising an error function. The algorithm solves the linear transformation and does not need to actually calculate the translation, rotation and reflection parameters.

In other words the process finds a transformation that minimises the distance between the GNSS position information and the node while keeping the UWB topology intact. The linear transformations are solvable using any known technique. Alternatively, the method may fix one node using its GNSS measurement and then match the topology to the most likely by finding the best fit of the topology to the remaining measurements.

Figure 5 illustrates the principle of fitting the topology to the GNSS measurements. The tags are shown as whole circles and the GNSS measurements of the tags are shown as clear circles. The estimation algorithm solves for the best fit of possible parameters that fits the topology to the GNSS measurements. Since the UWB are considered to be more accurate than the GNSS measurements, preferably the vertices and edges of the UWB topology do not change but the estimation algorithm determines the most likely approximation of that topology to the GNSS measurements.

In this way the process has used accurate UWB measurements to compensate for inaccurate GNSS measurements. Alternatively it can be considered that the GNSS measurements are used to fix the UWB topology to a fixed configuration.

In the exemplary process, a GNSS position of three nodes is needed. Although three points may define a plane, and a fourth point may define an up and down position, since the tags are located on the earth the algorithm is able to know which way is up and so only three GNSS measurements may be needed.

The topology may be fixed to a global coordinate system or a local coordinate system in the software algorithm or based on the GNSS solution used. For example, the locations of the tags may be fixed to the coordinate system of the GNSS devices or may be fixed to a coordinate system relative to a master tag for example when the accurate topology is known. Further, the coordinate system may be set by the sports field either through predetermined calculations, through an additional tag placed on or around the pitch or by mapping the GNSS data to known data for the sports field.

Figure 6 illustrates the above described process in the form of a flow diagram. At step 601 , the linear distances between tags is first calculated. At step 602, based on these linear distances a topology is determined. At step 603, GNSS measurements are retrieved for at least three of the tags. At step 604, an arrangement of the device topology is determined using an estimation algorithm such as a least-squares model. From this model, the positions of each tag can be identified at 605, preferably the positions being identified in a coordinate system.

Figure 7 illustrates a high-level block diagram. Four tags 21 , 22, 23 and 24 are shown. The tags may offload data to a server or other computing or processing device 71 after the sporting activity has finished or may be offloaded in real time wirelessly via an antenna 72 using any known radio technique. The data may optionally be stored in a database 73 for subsequent analysis 71.

It has been described above how a solution to addressing the deficiencies of existing local position systems that use ultra-wideband is to make mobile tags worn by players range to each other (and to tags inserted in the ball) and to construct, at each instant of time, a connectivity matrix which defines the topology of the system.

The topology fully defines the relative positions between all the players and the ball, however, the topology has no bearing on the orientation of the tags with respect to a coordinate system, i.e. the rugby field or the earth, as illustrated in Figure 4.

It was described above that the topology could be fixed, at each instant of time, relative to a coordinate system by locating at least three of the players with respect to this coordinate system. The method described is to use GNSS measurements from at least three of the tags. The at least three GNSS measurements can be used to calculate a linear transformation. The linear transform may be comprised of a translation, rotation and flip parameters, which fix the topology to the coordinate system in a way which minimises some error function, such as least squares, as illustrated in Figure 5.

It is proposed herein in an alternative or additional embodiment to augment or otherwise replace step 603 in Figure 6 with any position measurement which gives the tag’s location with respect to a known coordinate system (i.e. a global or local coordinate system).

In one example, the topology of mobile tags, which has been determined by identifying linear measurements between each tag and certain ones of the other tags, may be fixed using one or more additional ultra-wideband measurements to stationery and transitory but not necessarily permanent anchors.

Note that not all tags need to be ranged to the non-permanent anchors. Above it was proposed to use at least 3 GNSS measurements to fix the topology. In an example, the system may use at least 3 UWB anchors on tripods placed at known locations to locate at least three tags and hence fix the topology. The measurements between the non-permanent anchors and whichever tags are used to fix the topology may be performed using any ranging technique described above, such as and preferably, two-way ranging as also described above in detail.

These non-permanent anchors allow the system to offload data in real time and work indoors. This solution provides for accurate determination of the position of a moving electronic device and also the position of a mobile device which is not in communication with the fixed anchors.

High power consumption GNSS measurements need not be obtained and thus the tags can be made small, lightweight and long-lasting which is preferable when these tags are worn by athletes and is essential when they are inserted in a ball. Similarly, the system can be used where GNSS systems have been shown to be inaccurate such as indoors or in stadium environments (or other inferring environments).

The non-permanent anchors need not be located using survey equipment or located at height as they do not provide the accurate location measurements of each tag. Instead, they provide a reference to a global coordinate system to fix a topology where the topology is determined by locating each tag relative to its peers. Thus the positions of mobile tags can be located to a similarly high accuracy to a system which ranges to fixed anchors of an accurately known location but without the inconvenience of installation which is time-consuming, expensive and a source of irritation for potential users, since the players or equipment can only be tracked at the location where the anchors are installed, for example stadia or defined training pitches.

Moreover, accuracy problems caused by line of sight challenges are obviated since each tag is located relative to a plurality of proximal tags. Additionally, should those linear measurements and the mesh be slightly inaccurate due to the communication of the tags being through the human body for example, the accuracy can be improved by using an error approximation algorithm, such as least squares as above) and the linear measurements to the non-permanent anchors.

Following the example of Figure 8, which is similar to Figure 6 and illustrates the process in the form of a flow diagram, at step 601 , linear distances between tags are first calculated. As above, these need not be between all tags but must be of sufficient number to determine a mesh topology of tags (i.e. to at least two neighbours). At step 602, based on these linear distances a topology is determined. At step 803, linear measurements between at least three of the tags and auxiliary UWB anchors are retrieved. At step 604, an arrangement of the device topology is determined using an estimation algorithm such as a least- squares model. From this model, the positions of each tag can be identified at 605, preferably the positions being identified in a coordinate system.

An example implementation will now be described in the context of Figure 9. In this example there are four mobile tags, 21 , 22, 23 and 24. The left-hand side shows the true configuration while the right-hand side shows a valid configuration of the nodes with the distances unchanged. In this example, a topology of the mobile tags can be fixed to a global system by using at least three non-permanent UWB anchors 91 , 92, 93 which are at known locations. The valid configuration of nodes on the right-hand side is fixed by the three non- permanent anchors but the distances between the devices is unchanged.

The three anchors have been placed at known locations on the field corresponding to intersections of field lines demarking the area of play. These locations are (-50,35), (-50,70) and (0,70). For example, if the anchors are on tripods at a known height of, say, 1.6m then one can include their heights as (- 50,35, 1.6), (-50,70,1.6) and (0,70,1.6). These anchors determine their range to the certain ones of the tags and from those range measurements, using a known technique, determine the location of at least three of the players with respect to the global coordinate system (which in this example is centred at the intersection of the halfway line and the touch line as shown in Figure 9.)

The at least three player locations, determined from the non-permanent UWB anchors, can be used to calculate a linear transformation. That linear transformation may be comprised of a translation, rotation and flip, which fixes the topology to the global coordinate system in a way which minimises some error function. In the example given above, that error function may be for example, least squares, as illustrated in Figure 5.

Once the topology has been fixed, all the objects will be located in the global coordinate system. This is illustrated in Figure 9, where the tag 21 is now in its correct position despite not being located by the non-permanent UWB anchors.

At least one of these anchors may be used to offload data in real time as is illustrated in Figure 7 as item 72. Any one of the anchors may function as the master anchor as substantially described above. Similarly, one of the non- permanent anchors may function as the master clock as described above and function to propagate synchronising time signals throughout the mesh topology. One may consider the topology as a polyhedron in which the moving electronic devices or tags are the vertices and the linear distance measurements are the edges. In order to fix this polyhedron to a known configuration or arrangement, then 3 of the vertices need to be accurately located within a coordinate system. 3 are needed as it is possible to expect a particular solution as it is assumed that the tags are located relative to the ground. A 4th measurement will reduce symmetry but this is not necessary to accurately locate each of the tags and fix the polyhedron in a known coordinate space (a 4th measurement creates a negative solution).

Continuing this conceptualisation of the invention, the creation of a set of linear measurements creates the edges of a polyhedron. In the examples described above, all vertices can be fixed from three position measurements of three of vertices. If these position measurements are obtained by GNSS then the 3- dimentional nature of the GNSS measurement allows those three vertices to be fixed in position to a coordinate system. If ultra-wideband measurements are used (to obtain position measurements) to fix the position of the polyhedron in the coordinate system and convert the linear measurements to locations, then at least three ultra-wideband measurements between a vertex and a satellite anchor may be necessary. That is, each vertex (of the subset used to fix the topology) needs three unique ultra-wideband measurements to satellite anchors. Once these vertices (i.e. the subset) have been fixed in location, all the other ‘floating’ vertices can be located because their relative position in the polyhedron has already been accurately determined. In other words, fixed satellite anchors helps to locate at least three vertices of the polyhedron in order to fix the entire polyhedron.

Exemplary tags or devices will now be described. The tag has a transceiver, antenna and a microcontroller.

The devices may have at least one 9-axis IMU for collecting linear acceleration, angular acceleration and orientation data. The IMU component optionally embedded in the devices includes a combination of accelerometers, gyroscopes and magnetometers to report characteristics over time. Data from the IMU can be combined with the position data, which constitutes a form of sensor fusion, to increase system robustness. The IMU may also be used to calculate the angular velocity components, and therefore the revolutions per minute, of the ball or player device. The tag preferably comprises a control unit, for example comprising a microcontroller. The tag preferably includes a power supply to supply electrical power, for example to those sensors which require power to operate as well as to the components of the control unit. A lithium ion polymer battery may be used. The power supply may be provided on the same part of the sports equipment as the sensors but this is not essential.

The control unit and power supply may be provided in an electronics unit or device. Where the sports equipment is a rugby ball, the electronics unit could be located inside the ball. Where the sports equipment is body armour, the electronics unit may be located in a position corresponding to a point between the player’s shoulder blades, which is already common practice for devices which use GNSS enabled chipsets for location.

As described above, the exemplary system comprises a device to be located on the field on either the player or in some other equipment. The intelligence to perform this analysis may be spread across each node or may be performed centrally at a central server. The data may be logged in a database for subsequent retrieval and analysis.

Throughout the present description the terms analytic controller, central server and microcontroller are used to describe processing units which perform certain functions. It will be understood that the terms used are not essential. What may be essential is the functionality described. However, the functionality may often be performed by processing and control units located remotely, within the described entities or elsewhere in the system as appropriate.

Methods and processes described herein can be embodied as code (e.g., software code) and/or data. Such code and data can be stored on one or more computer-readable media, which may include any device or medium that can store code and/or data for use by a computer system. When a computer system reads and executes the code and/or data stored on a computer-readable medium, the computer system performs the methods and processes embodied as data structures and code stored within the computer-readable storage medium. In certain embodiments, one or more of the steps of the methods and processes described herein can be performed by a processor (e.g., a processor of a computer system or data storage system). It should be appreciated by those skilled in the art that computer-readable media include removable and non-removable structures/devices that can be used for storage of information, such as computer-readable instructions, data structures, program modules, and other data used by a computing system/environment. A computer-readable medium includes, but is not limited to, volatile memory such as random access memories (RAM, DRAM, SRAM); and non-volatile memory such as flash memory, various read-only-memories (ROM, PROM, EPROM, EEPROM), magnetic and ferromagnetic/ferroelectric memories (MRAM, FeRAM), and magnetic and optical storage devices (hard drives, magnetic tape, CDs, DVDs); network devices; or other media now known or later developed that is capable of storing computer-readable information/data. Computer-readable media should not be construed or interpreted to include any propagating signals.